ReferenceManual OrthoVista (English) 45

OrthoVista 4.
Reference Manual
Page 0
All rights to this publication are reserved. No part of this document may be reproduced, transmitted, transcribed, stored in a retrieval system, or translated into any language, in any form or by any means, without prior written permission from INPHO GmbH. The software described in this document is furnished under a license agreement. The software may be used or copied only in accordance with the terms of the agreement. It is against the law to copy this software on magnetic tape, disk, or any other medium for any purpose other than the licensees personal use. Copyright 2001, 2010 INPHO GmbH All rights reserved. OrthoVista Software Manual for OrthoVista Version 4.5 and higher INPHO GmbH reserves the right to make changes to this document and the software described herein at any time and without notice. INPHO GmbH make no warranty, express or implied, other than those contained in the terms and conditions of sale, and in no case is INPHO GmbH liable for more than the license fee or purchase price of this product. Sample data used in this manual provided courtesy of Digital Globe Incorporated. The sample imagery is of Castle Rock, Colorado U.S.A. The imagery has 5 m ground resolution and is in NAD83 UTM zone 13. Units are in meters.
Page 1
1
1.1 1.2 1.3
1.3.1 1.3.2 1.3.3
Introduction to OrthoVista......................................................... 1
Quick start................................................................................................... 1 What is OrthoVista? ................................................................................... 1 What can OrthoVista do for you? ............................................................. 1
Manual or semi-automatic adjustments................................................................. 2 Balancing and tilting .............................................................................................. 2 Mosaicking ............................................................................................................ 3
1.4 1.5
How does OrthoVista fit into your workflow? ......................................... 3 Will OrthoVista solve all your imagery problems? ................................. 3
2
2.1 2.2 2.3
2.3.1 2.3.2
Installing OrthoVista .................................................................. 5

What are the system requirements for running OrthoVista?................. 5 Linux installation........................................................................................ 5 Windows installation.................................................................................. 5
Windows registration ............................................................................................. 5 Linux registration ................................................................................................... 5
3
3.1 3.2
Getting Started with OrthoVista................................................ 7

Invoking OrthoVista ................................................................................... 7 Exiting OrthoVista ...................................................................................... 8
4
4.1
Basic Concepts........................................................................... 9
Main window ............................................................................................... 9
4.1.1 Show Histogram .................................................................................................. 10 4.1.2 To Front............................................................................................................... 10 4.1.3 Display Layers..................................................................................................... 11 4.1.4 Viewing Control ................................................................................................... 11 4.1.4.1 ....................................................................... Single-shot versus Continuous mode: 11
4.2 4.3
Shortcuts................................................................................................... 12 Plugins ...................................................................................................... 12
4.3.1 4.3.2 4.3.3
Image support plugins (image formats) ............................................................... 12 Geo-reference data support plugins (report formats) .......................................... 12 Image adjustment plugins (image processing) .................................................... 13
4.4 4.5 4.6 4.7

4.7.1 4.7.2 4.7.3
Rotated images......................................................................................... 13 File management ...................................................................................... 14 Project Dialog ........................................................................................... 15 Defining the processing area .................................................................. 16
Using an explicit tile definition file (standard file extension is txt) ........................ 17 Using a tile definition with simple parameters (standard file extension is tsp) ..... 17 Loading a tile definition file: ................................................................................. 20
4.8 4.9
4.9.1 4.9.2
Loading input image data........................................................................ 20 General Preferences ................................................................................ 22

Background pixels ............................................................................................... 25 Display mapping.................................................................................................. 25
4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18

4.18.1 4.18.2 4.18.3 4.18.4
Specifying a custom area ........................................................................ 26 Selecting one or a group of tiles: ........................................................... 27 Begin Processing ..................................................................................... 28 Clearing the defined processing area .................................................... 28 Displaying Exclusion areas..................................................................... 28 Displaying Automatically Generated Seams ......................................... 29 Display Water Areas................................................................................. 29 Clearing Vector data layer ....................................................................... 29 Image Commander ................................................................................... 29
Generation of Overviews..................................................................................... 30 Channel Assignment ........................................................................................... 30 Radiometrix ......................................................................................................... 31 View Image ......................................................................................................... 31
4.19
4.19.1 4.19.2 4.19.3
Radiometrix Editor ................................................................................... 31

Adjusting images with the Radiometrix Editor ..................................................... 31 Save/Reject changes of the Radiometrix Editor .................................................. 39 Radiometrix Editor: Background information ....................................................... 40
4.20 4.21
Color Picker .............................................................................................. 40 Status Dialog ............................................................................................ 40
5
5.1
5.1.1 5.1.2 5.1.3 5.1.4 5.1.5
Processing Options.................................................................. 42
Output options.......................................................................................... 43
Specifying the Output Directory........................................................................... 43 Meta Data Directory ............................................................................................ 43 Output Image format ........................................................................................... 44 Output Report format........................................................................................... 45 Save Background Information for Output Images................................................ 46
5.2
Adjustments options................................................................................ 46
5.2.1 Specifying radiometric adjustments for single images......................................... 46 5.2.2 Per-Image Selection............................................................................................ 49 5.2.3 Image group adjustment...................................................................................... 50 5.2.3.1 .................................................................................................. Global Tilting Adjustment 50 5.2.3.2 ............................................................................................. Contrast adjustment options 53 5.2.3.3 .......................................................................................................... Reflections Removal 55 5.2.4 Mosaic adjustment .............................................................................................. 58 5.2.5 Mosaic Hints........................................................................................................ 63
5.3
Output Selection....................................................................................... 64
5.3.1 Saving adjusted images ...................................................................................... 64 5.3.2 Rotate adjusted images before saving ................................................................ 64 5.3.3 Generate Seam Data (*.cld Files)........................................................................ 64 5.3.4 Save Vector Seams............................................................................................. 64 5.3.5 Save Vector Seams for each image .................................................................... 64 5.3.6 Seam simplification tolerance.............................................................................. 65 5.3.7 Save Mosaic Output ............................................................................................ 65 5.3.8 Options for Saving Adjusted Images and Mosaic Output .................................... 65 5.3.8.1 .................................................................................................... Internal Name of Output 65 5.3.8.2 ...............................................................................................................................Directory 65 5.3.8.3 ...............................................................................................................File Name Format 66 5.3.8.4 ..................................................................................................................................... Ratio 66 5.3.8.5 ....................................................... Number of channels and RGB component setting 66 5.3.8.6 .......................................................................................................... Channel assignment 66 5.3.8.7 ......................................................... Example for 4 Channel RGB and Infrared image 66
6
6.1 6.2
6.2.1
Advanced Information.............................................................. 69
Multi-Channel Image Support ................................................................. 69 Batch Mode Processing Capabilities ..................................................... 69
Examples ............................................................................................................ 70
6.3 6.4 6.5 6.6 6.7 6.8 6.9

6.9.1 6.9.2
Using user-defined vectors ..................................................................... 70 Radiometric Models ................................................................................. 71 Hot spot removal ...................................................................................... 73 Intensity Dodging ..................................................................................... 73 Coordinate reference ............................................................................... 74 Non constant pixel size, odd offset of orthophotos and tiles.............. 74 Processing large blocks with 32bit version of OrthoVista................... 76
Two step processing ........................................................................................... 76 Subdividing a block in sub-blocks........................................................................ 77
6.10 6.11
Processing large blocks with 64bit version of OrthoVista................... 77 Processing speed..................................................................................... 77
End User License Agreement ................................................. 79
Reference Manual OrthoVista
Introduction to OrthoVista
Heres a guide that refers you to sections of the manual that will get you started quickly with OrthoVista: If you already have OrthoVista installed and running, you can use icons found on the toolbar to guide you through the three simple steps to processing image data with OrthoVista: Load image data. This icon invokes the Choose Directory dialog so that you can specify the location of your input data and load the imagery for processing. Select processing area. This icon invokes the Custom Area dialog where you can define the image area to be processed. You can select all images or use your mouse to define a processing area. Process imagery. This icon invokes the Processing Options dialog. Simply specify an output directory, adjust any processing options if necessary and click the Process button.
1.1 Quick start
1.2 What is OrthoVista?

OrthoVista is a powerful software product that improves the quality, utility and value of ortho-rectified, digital image mosaics by performing a series of radiometric adjustments designed to match color and intensity across component images and producing seamless image mosaics.
1.3 What can OrthoVista do for you?

OrthoVista computes radiometric adjustments that compensate for visual effects such as hot spots, lens vignetting and mismatches between adjacent mosaic images. In addition it offers a powerful set of tools to manipulate the radiometry of single or group of images. Finally OrthoVista automates production of high-quality, photogrammetric orthophoto mosaics, providing the capability to define mosaic production quickly and easily on a project-wide basis.
Page 1
Figure 1:
Sample data set before OrthoVista processing.
1.3.1 Manual or semi-automatic adjustments

The built-in Radiometrix tool offers a set of tools to correct gradation, intensity, contrast, color and saturation, both manually and semi-automatically. In addition histograms and Color Picker can be used to check the changes.
1.3.2 Balancing and tilting

OrthoVista removes solar reflection hot spots and improves visual uniformity of most orthophotos by balancing the intensity and color variation across each frame. The software compensates for lens vignetting and various illumination effects by matching the color and intensity of adjacent input images in order to provide smooth and consistent radiometric image properties across all images. OrthoVista can correct color and intensity defects introduced during scanning or other processing as well. Further on water reflections can be eliminated and images can be interactively changed in color, brightness, contrast and saturation.
Page 2
Figure 2:
Sample data set after OrthoVista processing
1.3.3 Mosaicking
OrthoVista can be configured to generate seamless image mosaics from large numbers of individual orthophotos for use in Geographic Information System (GIS) and remote sensing applications.
1.4 How does OrthoVista fit into your workflow?

OrthoVista works with digital orthophotos in standard formats. The imagery produced by OrthoVista is immediately ready for delivery to clients or applications that use orthophoto mosaics. OrthoVista processes the individual orthophoto images and accompanying geodetic information that are typically produced by a soft-copy photogrammetric workstation or an orthophoto production system. Because processing can be performed in the background or after hours, OrthoVista can dramatically increase production efficiency.
1.5 Will OrthoVista solve all your imagery problems?

While OrthoVista does an excellent job of performing radiometric corrections for a majority of the imagery thats processed, there will always be extreme cases that automated image processing cannot address satisfactorily. Likewise, while OrthoVista will produce seamless mosaics most of the time, you may encounter problems with some cases involving exceptionally demanding imagery. Here are some specific issues that may be encountered during radiometric correction and mosaic production:
Page 3
Definition of background pixels. By default OrthoVista treats 0,0,0 as background. You can modify OrthoVistas interpretation of background values. See the Setting Preferences for information on manipulation of background pixel values. Large water bodies, islands and shorelines. Large water bodies can significantly impact OrthoVistas per-image balancing. In general, internal corrections (e.g., image balancing) are sensitive to water bodies, whereas the adjacent image adjustments ( e.g., global tilting) are not particularly affected. One option is to skip the internal image adjustments and only apply the tilting corrections (see the Modifying Image Processing Options section in Chapter 5). This will promote a more stable solution, but will leave hot spots in unbalanced imagery. Another option is to define an area to be excluded from processing as discussed in the Using user-defined vectors section in Chapter 5.2.1. Contrast adjustments. Severe contrast differences between adjacent images can result in a detectable visual difference between images. Radiometric adjustments have two methods (multiplicative and additive), which have different contrast effects. Experimenting with these methods may show that selecting one method may provide better results than the other for a given set of images. An effective approach is to manually adjust the contrast of input images that have extremely disparate contrast properties with the use of the OrthoVista Radiometrix Tool.
INPHO is committed to improving and enhancing OrthoVista. The software is continually addressing more of these demanding situations and will get better and better at handling problematic data; however, its important to note that results may not be perfect when dealing with particularly difficult imagery.
Page 4
Installing OrthoVista
To install OrthoVista, load your distribution CD-ROM or download an archive file from the OrthoVista download page on the INPHO or OrthoVista Direct website: http://www.inpho.de
2.1 What are the system requirements for running OrthoVista?

Processing time is determined mainly by CPU speed and even more important disk access speed. OrthoVista requires 2 GB RAM on a 32 bit system and and a 64bit system a minimum of 6GB RAM is recommended. Always try to run OrthoVista with image data on fast, local disk storage device. Should you use network drives, we strongly suggest to always map the network drives, use a 1 Gbit network or even faster otherwise the processing performance will be severely affected.
2.2 Linux installation

You will find a detailed description in the InstallationManual_inphoSoftware_(Englisch).
2.3 Windows installation

Just start the setup.exe and follow the installation instructions.
2.3.1 Windows registration

OrthoVista uses a hardware dongle to protect the software. You will get the dongle when you purchase OrthoVista. It is also possible to run OrthoVista with a server license where the dongle is managed and located on a server accessible via the network.
2.3.2 Linux registration

Starting with version 4.3 of OrthoVista, the Linux version of our software uses hardware licensing via WIBU CodeMeter. This requires that the CodeMeter software is installed and running on any computer that should run the software, even if the license is acquired over the network. The latest version of the CodeMeter runtime is always available under: www.codemeter.com For convenience, the current version is bundled with this software. Installation files for Debian based systems (Debian, Ubuntu, ...) can be found under
Page 5
opt/inpho/deb-packages, those for rpm based systems (SuSE, RedHat, ...) can be found under /opt/inpho/rpm-packages. Both directories contain subdirectories for 32bit systems (i386) and 64bit systems (amd64 or x86_64). Please pick the appropriate installer package for your system and install it as usual. Special note on Debian 4.0 (Etch) on 64bit systems: At least version 4.0r3 does not provide the package ia32-libs-gtk, which is required by the CodeMeter package. However, the package is available as a backport and can be found under /opt/inpho/deb-packages/amd64. Other distributions based on Debian Etch (like Ubuntu) do provide this package, so you might try to install it using apt-get first. After installation, you might delete the directories /opt/inpho/deb-packages and /opt/inpho/rpm-packages.
Page 6
Getting Started with OrthoVista

To start OrthoVista, use one of the following instructions: On Windows systems, double-click the OrthoVista icon. On Linux systems, type orthovista (or the alias youve assigned) at any command prompt. (Remember that Linux is case-sensitive.)
3.1 Invoking OrthoVista
Figure 3:
The OrthoVista Main Window
Once started, OrthoVista displays its main window. Using the software generally involves three simple steps that can be initiated using the icons below available in the main windows toolbar: Load image data. This icon invokes the Choose Directory dialog so that you can specify the location of your input data and load the imagery for processing. Select processing area. This icon invokes the Custom Area dialog where you can define the image area to be processed. You can select all images or use your mouse to define a processing area. Process imagery. This icon invokes the Processing Options dialog. Simply specify an output directory, adjust any processing options if necessary and click the Process button. Refer to the OrthoVista Tutorial for more help getting the software up and running quickly.
Page 7
To start using OrthoVista, you need geo-referenced image data. This data must consist of individual orthophoto images and accompanying geodetic reference information. OrthoVista works with a variety of georeference data generated by common mapping, remote sensing, photogrammetry and image processing software. Orthophoto data typically are produced using a soft-copy photogrammetric workstation, an orthophoto production system and/or various remote sensing software packages.
3.2 Exiting OrthoVista

To quit OrthoVista: Select Quit from the File menu.
Page 8
Basic Concepts
OrthoVista provides a number of tools that controls the way in which your imagery is displayed in the main window.
4.1 Main window
Figure 4:
Main window
Once youve loaded your image data, you can manipulate the display using either: The icons displayed on the left side of the main window. The Display Layers pop-up menu.
Figure 5:
Display Layer Popup Menu
To activate the Display Layers pop-up menu, right-click in the main window and hold the mouse button down for at least 1 second.
Page 9
4.1.1 Show Histogram

To display histograms of single images, right-click on a certain image and hold the mouse button down for at least 1 second. A pop-up menu allows you then to open the histogram of this image.
Figure 6:
Image histogram
4.1.2 To Front
If images overlap each other at a certain position, OrthoVista displays the last loaded image on top of all the others. To change the sequence of images, rightclick on a certain position and hold the mouse button down for at least 1 second. A pop-up menu allows you then to select a certain image which will be then on top of all others.
Figure 7:
Selection of an image to be on top
Page 10
4.1.3 Display Layers

Each option can be turned on or off independently. Tile Borders: Displays the tile definition borders, if any are defined. Ortho Borders: Displays the borders of the input images in red. Images: Displays the input images within the ortho borders. Output Areas: Displays the physical area (Tiles) to be processed.
4.1.4 Viewing Control

The View Control Tool Bar contains a number of tool buttons used to change the area displayed in the Main Window.
Zoom window mode: This is the standard mode for zooming. By clicking and holding the left mouse button, drag a rectangle over the desired view area. Release the left mouse button, and the view display will zoom to the defined area. A single right click will zoom back to the previous zoom level. Note: If the zooming rectangle is too small, no zooming will take place. To zoom into a tiny rectangle, you might have to use two subsequent zooms. Panning mode: When panning is active, you can grab the image shown in the Display Area and drag it to a new position. This is similar to using the scrollbars of the Display Area. See also 4.1.4.1. Zoom-in mode: While in this mode, each click with the left mouse button will recenter at that location and also zoom in by a factor of two. A single right click will zoom back to the previous zoom level. See also 4.1.4.1. Zoom-out mode: In this mode, each click with the left mouse button re-center at that location and also zoom out by a factor of two. A single right click will zoom back to the previous zoom level. See also 4.1.4.1. Zoom reset button: This button resets the zoom stack and shows the whole project in a fit-to-view zoom level. 4.1.4.1 Single-shot versus Continuous mode: The panning and the zoom in/out buttons operate in two different modes. A single click on the button involves a single-shot operation. Once this operation is activated the view control falls automatically back to the zoom-window mode. Double clicking these buttons enters the modes permanently until a different zoom/pan button or the zoom-window button is pressed.
Page 11
4.2 Shortcuts
Zooming, panning and window sliders can be also controlled by using hard-key shortcuts. The following shortcuts are supported: Shortcut + F C Arrow Keys Function Zoom in Zoom out Full Screen or Reset Display Center Position Move Slider up/down/left/right
4.3 Plugins
OrthoVista supports an innovative architecture that packages specific product functionality into modules called plugins. The advantage of using a modular architecture is that OrthoVista only loads the modules required to perform the processing jobs that you specify resulting in faster, more efficient processing. Plugins support third-party development and product customization. They encapsulate product features such as image format support and specific image processing capabilities. This allows independent development and delivery of schedule-critical and/or proprietary processing modules. Plugins are grouped into several categories; you can get details on each specific plugin from OrthoVistas Info menu.
4.3.1 Image support plugins (image formats)

OrthoVista accepts a variety of image formats: TIFF (scanline or tiled) and GeoTIFF BIP/BIL/BSQ To determine which specific image formats are supported by your installation of OrthoVista ( e.g., list the image support plugins that are currently installed), select About Plugins from the Info menu and select Image Support Plugins. A dialog will display the plugins currently available and provide detailed technical information about image formats and their specifications.
4.3.2 Geo-reference data support plugins (report formats)

These plugins are used to read and write image georeference information. GeoTIFF .tif format TIFF World .tfw and .tifw formats Vision Softplotter .rpt format Zeiss Phodis .inp format ER Mapper raster file format .ers GeoTIFF Note: OrthoVista handles GeoTIFF files that contain the image and header data in the same file, and supports the following tags in GeoTIFF format: ModelTiepointTag ModelPixelScaleTag
Page 12
ModelTransformationTag Other tags that describe projections as well as extensions that are not supported will be carried through the process and written in the output files without alteration. More information concerning GeoTIFF can be obtained from OrthoVistas Info menu.
4.3.3 Image adjustment plugins (image processing)

Many of OrthoVistas processing capabilities are implemented as plugins: hot spot removal intensity dodging global tilting adjustment reflections removal image rotation plain mosaicking sheet centered mosaicking adaptive feathering seam applicator Check the Info menu to determine the plugins currently installed. Custom plugins may also be developed to meet an individual customers specific image processing requirements.
4.4 Rotated images

OrthoVista can accommodate rotated input images. It can also rotate the output image so that mosaics are generated with arbitrary pixel row/column azimuths. If all input images are not aligned, then image sampling is required. Consequently, processing of rotated images can be substantially slower than processing nonrotated imagery. OrthoVista automatically detects input image rotation when it reads the input orthophoto georeference data. To rotate an output mosaic: Click the canvas orientation button in the lower-left corner of the main window.
Figure 8: OrthoVista displays the Orientation dialog. Select Degrees, Gradians or Radians from the Angle Units drop-down box. Click in any one of the directional data entry boxes (e.g., North, South, East and West) and type the angle of rotation you would like to apply to the input images.
Page 13
Note: Processing areas and output data are always parallel to the main window, and all output areas must have the same rotation for a given processing run. When you rotate the display, previously defined processing areas are automatically cleared and you must redefine them. OrthoVista rotates both the input images and the tiles (if you defined them) by the angle you specify. The main window stays aligned with the output mosaic.
4.5 File management
Create a new project: Creates a new project and clears data that was previously loaded in OrthoVista. Open an existing project: Displays a file selection dialog where you can select and load an existing project. Save the current project: The project itself and the selected processing area respectively tile definitions are written to disk. Imports a Configuration File. The configuration is imported and written into the orthovista.cfg file located on the directory C:\Documents and settings\All Users\Application Data\Inpho\OrthoVista4\Settings (on Windows 2000 and Windows XP) respectively C:\ProgramData\Inpho\OrthoVista4\Settings (on Windows Vista/7). Saves the parameter settings into a user defined Configuration File. Printing the display OrthoVista allows you to print what appears in the main window. You can output to printers and/or postscript files. To print the current display: Select Print Display from the File menu. The printed output represents what is on the screen, so zooming and scrolling the main window will change what is printed. For best results, resize the main window to match the proportions of the page to which you are printing.
Page 14
4.6 Project Dialog
The project dialog allows parallel to the handling in the main window the following functions: Loading orthophotos by file or directory Like in the main menu, the project dialog allows to load either single or a group of files or a whole directory. See also chapter 4.8. Unload selected images allows to unload images after they are selected first in the project dialog. Change file path allows to change the file path of images after they are selected first in the project dialog. This function is needed after having loaded a project file and the file path of the images in the project file is no more correct.
Page 15
To change the path, select images first, press the Change Path button select a directory. The new directory will then be automatically applied.
to
The meta data directory can be changed as well by using the Browse button. Activating/Deactivating images allows to activate/deactivate images after they are selected first in the project dialog. Only activated images are displayed in the main menu and processed. Select/Unselect images allows to select/unselect images for the Radiometrix Editor after they are selected first in the project dialog. Load tiles allows to load tiles from a tile definition file. See also Chapter 4.7. Unload tiles allows to unload tiles after they are selected first in the project dialog. Mark/unmark tiles for processing allows to mark/unmark tiles for processing after they are selected first in the project dialog. Change edited status of tiles after they are selected first in the project dialog. The Edited status is automatically modified by the Seam Editor when seam lines influencing a certain tile are modified. Once seams are modified with the Seam Editor the changes have to be applied with a Seam Applicator run. In order to avoid that all tiles have to be reprocessed, the Edited status helps to check which tile has to be re-processed. The Tiles list can be sorted for Edited and non-Edited tiles and once sorted the tiles can be selected and marked for processing. If further on seams are modified it is helpful to set the tile status back to not-Edited before starting a new Editing.
4.7 Defining the processing area

For your convenience, you can define specific areas of the input image data for processing.
Page 16
Each area defines an output file. Radiometric corrections are only computed for the areas that are defined by the collection of all processing areas. OrthoVistas processing does not include images whose area is not included in a defined processing area. A common first step in production operations is to create a tile definition file. Often, the output tile definition is created during flight planning or project management activities. OrthoVista can read tile definition files and use them to define output mosaic boundaries. The following sections describe alternative techniques to define the image processing areas.
4.7.1 Using an explicit tile definition file (standard file extension is txt)
To create this form of tile definition file, use any text editor, spreadsheet, or processing script to write an ASCII .txt file in the format below. OrthoVista ignores lines where the second column has double quotes ( e.g., ). The file consists of data in five columns: <tile-id> ascii string <northwest-X-coordinate> floating point number <northwest-Y-coordinate> floating point number <southeast-X-coordinate> floating point number <southeast-Y-coordinate> floating point number A sample file might look like this: "TileID" "NWx" "NWy" "SEx" SEy" "tile-A1" 470000 4510000 480000 4500000 "tile-A2" 480000 4510000 490000 4500000 : : These files are read by a tile definition plugin. Select About Plugins from the Info menu to determine the tile definition plugins supported by OrthoVista and to obtain detailed specifications for each format.
4.7.2 Using a tile definition with simple parameters (standard file extension is tsp)
To create this form of a tile definition file, use any text editor, spreadsheet, or processing script to write an ASCII .tsp file in the format below. File is a 'keyword value(s)' format with content: NOTE: Keywords INCLUDE the ':'(colon) character!
-----------#Everything after a '#' character is comment TileCorner: x0 y0 TileSize: dX dY TileSkip: dX dY TilesToWest: nXW TilesToEast: nXE TilesToNorth: nYN TilesToSouth: nYS
Page 17
TileNameFormat: fmt # end of file -----------Where: : x0, y0 : are East/North of any exact corner : sizeX, sizeY : are X(E-W) and Y(N-S) tile dimension : dX, dY : are X(E-W) and Y(N-S) distance between tiles : nXW : number of tiles to west of corner : nXE : number of tiles to east of corner : nYN : number of tiles to north of corner : nYS : number of tiles to south of corner : fmt : format for the output file names (see below) TileSkip is optional and defaults to TileSize and can be used to define tiles, which overlap each other. TileNameFormat is optional. The default is a row and column based numbering that guarantees unique tile numbers. Example: #----Start of File # place tiles on even 10k grid # NOTE: all units are 'world/ortho' units TileCorner: 120000 1320000 TileSize: 10000 10000 # assume corner is upper left (N/W) of project area # and that project area is covered by 5 by 7 tiles # but we also want to pad one extra column to the west # NOTE: units are 'number of tiles' TilesToWest: 1 # pad one column west of corner TilesToEast: 5 TilesToNorth: 0 # no rows above corner TilesToSouth: 7 TileNameFormat: t%03.3ulx%03.3uly #----End of File NOTE: numbers are expressed in decimal notation (e.g. the decimal is represented by a '.'(dot) character. Do _not_ use ','(comma) characters!
Format String Description: A format string can be an arbitrary string that contains placeholders for tile specific information. The placeholder structure is as follows: Structure Option 1: a '%' sign an optional '0' (leadingZeroIndicator) an optional number (fieldWidth) the character 'e' or 'n' (eastNorthIndicator)
Page 18
Structure Option 2: a '%' sign an optional '0' (leadingZeroIndicator) an optional number (fieldWidth) an optional '.' followed by a number (cutLength) 'ul', 'ur', 'll', 'lr', or 'c' (cornerSpecifier) the character 'x' or 'y' (rightUpIndicator)
The first structure is to include the horizontal ('e') or vertical ('n') tile number. It is printed using up to 'fieldWidth' digits. If there is a 'leadingZeroIndicator', there are exactly 'fieldWidth' digits; missing digits are padded with 0. Given the horizontal tile number 13, '%3e' gives '13', while '%05e' gives '00013'. The second structure is used to include the X ('x') or Y ('y') coordinate of a tile corner or the center in the tile name. The 'cornerSpecifier' is 'ul' for the upper left corner, 'ur' for the upper right corner, 'll' for the lower left corner, 'lr' for the lower right corner, and 'c' for the tile center. 'fieldWidth' - as above - specifies the maximum number of digits in the output, and 'leadingZeroIndicator' also behaves as described above. In addition, 'cutLength' gives the number of digits to be truncated from the right after rounding the coordinate value. Example: Assume the x coordinate of the upper left corner is 167874.738. First this is rounded to integer, giving 167874. A placeholder structure '%ulx' will thus result in '167874', '%8ulx' gives '167874', and '%08ulx' results in '00167874'. If 'fieldWidth' is less than the number of digits, digits are truncated at the left. For our example, '%4ulx' results in '7874'. To truncate from the right, 'cutLength' can be used. Thus '%.2ulx' gives '1678'. A combination might be '%2.3ulx', which results in three digits cut from the right, then taken the rightmost 2 digits: '67'. One more example: assume your coordinates are in meters, and you want a threedigit kilometer value of the lower left corner. The tile name must have the prefix 't_' and the X and Y values have to be marked with 'E' and 'N'. The format string for this tile specification is 't_%3.3llxE%3.3llyN'. For a lower left corner of (7521344.0 ; 13557412.0) the tile name will be 't_521E557N'. Example with TileSkip: TileCorner: 10000 TileSize: 1000 TileSkip: 900 TilesToWest: 1 TilesToEast: 1 TilesToNorth: 1 TilesToSouth: 1
10000 1000 900
Above settings lead to following tile corner coordinates:
Page 19
Figure 9: Example with TileSkip definition. The corners of the tiles are computed as follows: The given TileCorner defines the upper left corner of one tile (shaded tile). There are no more tiles to be generated to the south east as the one already defined To the north east, north west and south west the TileSkip value of 900 units is added and then an area of 1000 by 1000 units is defined. See tiles 11, 12, 22 so that the tiles overlap each other by 100 units. TileSkip defines the skipping distance from upper left corner to the next corner in x and y. At this corner then an area of TileSize dx, dy units is defined.
4.7.3 Loading a tile definition file:

Select Load Tile Definition from the Setup menu. OrthoVista displays the Open dialog. Locate and select your tile definition file and click OK. OrthoVista displays the tiles youve defined in blue. Note: TileSpec files must have the file extension tsp and explicit tile definition files must have the file extension txt.
4.8 Loading input image data

To process your input image data, you need to specify the location of the inputimage data along with the georeference data (e.g. tfw). The image-data files and the georeference data files must be in the same directory. Note that all the georeference data files found in the directory will be utilized. If you dont want to process a specific image, move the georeference data
Page 20
file to another directory before proceeding. Alternatively, you can load individual image files rather than all of the files in a directory. After loading one set of images, and you decide to add a second (new) set of images, you have the option of combining the new set of images to the existing data set or using the new set of images to replace the previous set. Note: If you are using GeoTIFF images with accompanying geodetic reference data files (e.g. TIFF world files) the OrthoVista status window will inform you that there are 2 geo-reference information. In this case, the TIFF World files (tfw) are overruling the geotiff information. To load input imagery from a directory: Select Load Orthos by Directory from the Setup menu or press the corresponding icon buttons. OrthoVista displays the Find Directory dialog. Select the directory containing the input images and click OK. OrthoVista loads the imagery into the main window and displays the boundaries of each individual input file in red. To load input imagery by file: Select Load Orthos by File from the Setup menu or press the corresponding icon button. Select either tfw, ads or tif (in case of geotiff images) files. OrthoVista displays the Specify Input Files dialog. Select the input image(s) and click OK. You can load each image individually or select multiple images. If you select multiple images, OrthoVista checks for georeference information (e.g. tfw files) and loads only these files. Once youve loaded your input imagery, you can modify the display as explained in the previous section. In case of ADS Pushbroom files select the ADS files to load the images. This will then load automatically all files belonging to an ADS file.
Page 21
4.9 General Preferences

By Selecting the Set General Preferences function the following Global settings window is started.
Figure 10:
General Preferences Dialog
Standard Cache Size This parameter defines a cache size, which is used to cache meta data for computation purposes. For 32bit operating systems, we suggest using the default value of 128 MB for normal images and 256 MB for large images like the ADS40 images or when the images do have a large overlap. For 64bit operating systems, the cache size can be set to 1024 when having at least 6GB RAM in the computer. Note: Beside the given Cache Size, OrthoVista needs more memory to process data. Especially when Save Vector Seams is activated. If too much cache size is defined with larger projects the physical memory will be reached. The operating system will then either kill OrthoVista or the computer uses the virtual memory, which slows down processing considerably.
Page 22
According to our testing, a larger cache size does not necessarily speed up processing but can enhance the reaction time in the Radiometrix Editor or if rgn files keeping Background Pixel information are large. Seam Editor Cache Size This parameter defines a cache size, which is used to cache meta data and images for Seam editing. For 32bit operating systems, we suggest using the default value of 384 MB. For 64bit operating systems, the cache size can be set to 1024 when having at least 4GB RAM in the computer. Log File This parameter is for problem tracking only. Should you have a problem with the Seam Editor then it might be helpful to define a log file and enable the Verbose switch. When done try to reproduce the problem and send the inpho support team the log file.
Display Background Color This parameter defines the background color of the OrthoVista main window area. Enable Parallel Processing OrthoVista is writing the images with multiple processes in parallel when this option is enabled. When enabled the number of parallel processes can be selected. Maximum number is 2. Do not select more processes then cores are available on your computer. Should you have only one core (CPU) available when also suggest to enable this option but to select 1 for the number of sub-processes. This job will run faster than without having enabled the parallel processing. Dependent on the file IO speed there might not be much improvement in speed between 2 sub-processes and more. Only by improving file IO 3 or even 4 sub-processes will allow faster processing. Overview Generation Overviews or a Full Set of Overviews are down sampled (minified) images of the original input image. They are stored in separate Files in the same directory as the input images. They have the file extension .pyr. The overview files are very important for a fast display of the images and for fast processing. Therefore we strongly suggest generating overviews if they do not exist. If the parameter No Overviews is activated OrthoVista will not create Overviews. If the parameter Single Overviews is activated, then OrthoVista will generate one single Overview per image (if it does not already exists) with an extension of about 1024x1024 pixels. If the parameter Full Set of Overviews is activated, then OrthoVista will generate several overviews with a downsampling rate factor of 2 from overview to overview. A full set of overviews allow a much faster zooming into the images, and is especially helpful if you use the images in the Seam Editor.
Page 23
If Delay Overview Generation is activated, then OrthoVista computes the Overviews only if they are opened for display or accessed for computation. Otherwise OrthoVista will start the overview generation in a multi-threaded mode as soon the images are loaded. Note: No computation of overviews for an image is done if either one or several overviews already exist. Enable Background Checking To enable automatic detection of background pixels, click on the Enable Background Checking checkbox. If your images have no background pixels you can disable this parameter. If disabled all pixels will be treated as valid image data, and the processing is faster. Minimum Non-Background Image Data Value Maximum Non Background Image Data Value To set the pixel values that will be considered background, change the minimum and maximum values in the Image Data Values box. If your images have black borders, set a range from 1 to 255 for 8 bit and 1 to 65535 for 12/16 bit images. If they have white borders, set a range from 0 to 254 for 8 bit and 0 to 65534 for 12/16 bit images. If you do not have black or white borders, disable Background Checking. You can define the Min/Max values independent on the input or output image resolution (8, 12 or 16 bit). The 16bit options give higher resolution steps. Note: If you define the Min/Max range from 1 to 254, you are defining both black and white as background information. In this case OrthoVista tries to fill such background colors with valid image data from overlaying images. If such areas cannot be filled from overlaying images, OrthoVista uses the parameter Output Background Color to fill the areas. It is possible that a white area is transformed to black or vice versa. Output Background Color This parameter defines the color to be used to fill Background Color areas if the areas cant be filled by valid image data. Display Mapping To control which image data are displayed in a specific display channel, change the values in the Display Mapping boxes. See also chapter 4.9.2. Note: Changing the settings for Min/Max Image data values and Background Checking invalidates all meta data files generated with the previous settings: This includes: *.rdx Radiometrix data files
Page 24
*.rgn Boundary Region files *.bal Image Balancing files *.tlt Group Adjustment files If you apply the changes, you have to delete all previously generated meta data files in the meta data directory and the *.rdx files manually. In addition you have to restart OrthoVista before the changes are effective.
4.9.1 Background pixels

Images are considered to have two pixel types consisting of either image pixels or background pixels. A common situation in which this occurs is when a rotated image is ortho-rectified. For example, if a square image is aligned 45-degrees from North and is rectified into a North-South coordinate system, the resulting data file will consist of a diamond-shaped area of image pixels surrounded by triangular areas of background pixels. Because background areas are commonly produced during ortho-rectification operations, OrthoVista provides several options for handling them. A significant feature of OrthoVista is the ability to automatically detect the border between the background pixels and the image pixels. The distinction between image and background pixels is controlled by the Image Data Values minimum and maximum values. By default, background pixels are defined as pixels, which have an intensity value of 0 (0,0,0 for black) or 255 (255,255,255 for white). These values can be adjusted higher or lower respectively to determine what will be considered background pixels. If the pixel intensity value is greater than or equal to the minimum and is also less than or equal to the maximum, then this pixel is considered to be a valid image pixel. If the intensity is outside of this range (and is connected by other background pixels that touch the edge of the data file), the pixel is classified as a background pixel. In mosaicking, the pixels will be treated as background if they are outside the edge of the image data. However, in image balancing, all values that are background will be ignored no matter where they are located.
4.9.2 Display mapping

OrthoVista allows handling multi-channel images. But a color monitor can display only 3 channels as RGB images. For a multi-channel image it is therefore necessary to define which channel shall be combined to show up as Red, Green and Blue. See also chapter 4.18.2 for a description of the RGB channel assignment. Note that grayscale images are always displayed as grayscale independent of the display mapping settings (e.g., the 0 image channel is displayed equally in the display R, G and B channels). The image channels are denoted by the names Red, Green, Blue and if more channels are available with a number starting with 0. For example, a grayscale image has only channel 0. A typical color image has three channels: Red, Green
Page 25
and Blue. The display has three-color channels: red, green and blue commonly identified as R, G and B, respectively. You can use these setting to view individual image channels. For example, if you want to view only image channel 0 (typically the red channel of a color image) as a grayscale image, you can set all display channels to this same image channel.
4.10 Specifying a custom area

If you dont have a tile definition file, you can use this function to define tiles for processing. To specify a custom area: Click Select Area
Figure 11:
Custom Area dialog
Click in the Tile Id box and type a unique name for the new tile you want to define (e.g., area01). Be sure to use different names for each Tile ID you define. OrthoVista will warn you if you try to use the same Tile ID twice. To define the custom area, do one of the following: Click and drag a rectangle in the main window. Type exact coordinates in the data entry boxes and click the Add Tile button. Click the Select All button. OrthoVista displays the custom area youve defined (green shaded). The new tile will be stored in the project as an area to be processed. You can select multiple areas for processing at the same time, just be sure to use different names for each Tile ID. To deselect the area, left-click on the main window while the Custom Area dialog is still open.
Page 26
4.11 Selecting one or a group of tiles:

After having loaded tiles with Load Tile definition (see chapter 4.6) it is necessary to define which tile shall be processed. This definition is done with the Tile Selection. To select an area of predefined tiles, you can select individual tiles or groups of tiles. Choose Select Tiles.
Figure 12: Select Tiles dialog. Select From Layer drop-down box: Choose Tile Borders to define a processing area by clicking on the blue tile borders (assuming youve loaded a tile definition file). Choose Ortho Borders to define a processing area by clicking on the red image borders to select all tiles that overlay the image. Selection Mode. In Tile Borders mode, you can define the processing area by selecting individual tiles (by their tile or orthophoto borders) or by dragging to select multiple tiles. If you set the Selection Mode to Single Tile, simply click each tile in the main window that will become part of the processing area. You can select one or more tiles. If you click on a tile again it will be deselected from the processing area. If you set the Selection Mode to Select Area, you can use your mouse to click and drag a selection rectangle around a group of tiles to define the processing area. Dragging a rectangle across tiles that are already selected will deselect them from the processing area.
Page 27
Figure 13:
Tile selection
The tile selection display can be turned on or off using the bottom icon found on the left side of the main window. To deselect one or a group of tiles: In single tile Selection Mode, click on the specific tile to be deselected. In select area Selection Mode, drag through the selected tiles again to deselect them. (Note that when the Select Tiles dialog is open, this operation replaces the zoom operation in the main window.)
4.12 Begin Processing

The begin processing button starts the setup window for processing options, and allows starting the process after having defined some of the parameter options. See chapter 5 for a detailed description.
4.13 Clearing the defined processing area

This function clears all previously defined processing areas and tile selections. It should be used if you want to restart with new definitions of processing areas or tile selections.
4.14 Displaying Exclusion areas

This function allows loading and displaying DXF or ArcShape files containing Polygons, closed Polylines/Arcs or DXF files with polygons stored in a layer with
Page 28
the capital letter named EXCLUSION. The purpose of it is to check the layer data.
4.15 Displaying Automatically Generated Seams

The button Load Seam Area displays a DXF file, which contains the seam polygons, generated by a previous OrthoVista run. If you press this button, OrthoVista opens a file selection box, which allows selecting a DXF file. Note: After selecting the file, OrthoVista reads all Layer data called SIMPLIFIED and displays it on top of the image data.
4.16 Display Water Areas

The button Load Water Areas allows displaying either DXF or ArcShape files containing vector information to be used for the Reflections Removal function. Note: In case of DXF files only layer data is considered with the name REGION.
4.17 Clearing Vector data layer

Press the button Clear Region to clear the vector layer data.
4.18 Image Commander

The Image Commander allows to Generate overviews for the images Assign which channels of an image contains the RGB channels View single images The Image Commander is a tool available for all ApplicationsMaster applications and is now also integrated in OrthoVista. The main purpose for OrthoVista is the assignment of the RGB channels.
Figure 14:
Image Commander window

Page 29
4.18.1 Generation of Overviews

Select first the images for which overviews shall be generated and then press the button Generate Overviews.
Figure 15: Generate Overview option dialog Define your options for the overview generation and press then the Start button. Note: The Schedule Task option is at time not supported in OrthoVista.
4.18.2 Channel Assignment

The channel assignment allows you to define if and which channel of your image contains the Red, Green and Blue bands. This assignment allows you later on to address the channel with the names Red, Green and Blue. Select first the images for which you would like to assign the RGB channels and then press the button Channel assignment.
Figure gure 16: RGB Channel Assignment option dialog

Page 30
Define now if your selected images contain RGB information and in which channels they are.
4.18.3 Radiometrix
This option allows you to start the Radiometrix Tool. See chapter 4.19 for a detailed description on the tool. The difference here is that the Image Commander lets you select the images for which you would like to use the Radiometrix Tool.
4.18.4 View Image

The image viewer allows selecting a certain image and start then the Image Viewer, which is also a standard viewer developed for the ApplicationsMaster environment.
4.19 Radiometrix Editor

During normal processing, OrthoVista matches the colors, contrast and intensity from one image to the next throughout the processing areas. If the input images are fairly uniform in color, contrast or intensity, you probably wont need to use the Radiometrix Editor. However, if some images were flown at different dates or if the film or scanner settings varied for different images (or groups of images), you can use the Radiometrix Editor to make corrections to the selected images. The Radiometrix Editor also can be used to stretch the histogram of 16 bit images that appear dark or almost black. The reason for the dark images is that usually the digital sensors only are able to record 11, 12 or 14 bit, but store the image in a 16 bit format.
4.19.1 Adjusting images with the Radiometrix Editor

By starting the Radiometrix Editor the first time OrthoVista computes for all images the min/max and mean values in color/intensity/contrast and saturation. These values are then displayed in different tabs.
Page 31
Figure 17:
Radiometrix Editor
Zoom in/out
Can be done by pointing with the cursor on a position and pressing the + or - keys on the keyboard. If Modify is selected zooming can be also done by dragging a rectangle with the left mouse button. Pressing the right mouse button zooms to the previous zoom level.
Selecting/Unselecting images
The Selection/Unselection of images can be done either in the project dialog or in the main window. Selection/Unselection in the Project Dialog is done by selecting first images of the image list and pressing then the select or unselect buttons. Selection/Unselection in the Main Window is done by pressing the Select/Unselect button in the Radiometrix Editor and then clicking with the left mouse button on an image or by dragging a rectangle. All images can be selected or unselected by pressing in the Radiometrix Editor the Select All or Unselect All buttons.
Color Space
Page 32
The color space shows the average color of each individual image. The red dot in the center of the color editor represents the center of the color space (perfect gray).
Intensity/Contrast Space
The intensity/contrast space shows the average intensity and contrast of each individual image. The intensity changes from left to right (dark to bright) and contrast changes from top to bottom (low contrast to high contrast). The red line represents the mean intensity.
Saturation Space
The saturation of the images is displayed in the so-called YUV color model. In the YUV color model the color is decomposed into three components called Y - luminance U - Cb: Chroma channel, U axis, blue component V Cr: Chroma channel, V axis, red component The range of the U and V values is from 0.5 to +0.5 U = V = 0.0 is gray, +/- 0.5 are the extreme color values. Each dot represents the average of the U/V values of an image. The red line represents the centerline of the UV range. To make adjustments, you should zoom into the dot display area since very small changes have a large impact on the saturation which is highly important to image quality. Changing the position of an image along/parallel to the red line, enhances the saturation for all colors. Moving the dot away or closer to the line changes the saturation for a specific color only.
Automatic Gradation and Intensity correction

This function automatically manipulates the histogram of images, basically to enhance brightness and color. To do this, the histograms of the images are analyzed and the brightest and darkest values are determined. Then the brightest value is defined as white and the darkest as black, and all the other values are proportionally spread.
Page 33
The automatic correction helps to get good results if the full image color range is using the complete histogram, e.g. a 12 bit image stored in a 16 bit file or if the image has a color cast. However the interactive correction can be more precise. To use the function select first the images to modify (See further information below). Then press the Run Auto-Adjustment button. The Automatic Gradation correction is preferred for RGB images, as also slight color casts can be corrected. The Automatic Intensity correction is preferred for IR (infrared) imagery, as here, color changes are not wanted.
Channels:
Common Adjustment Based On All Channels Analyzes the histogram of each color band individually. The darkest value of the color bands altogether and the brightest value of the color bans altogether is used as a reference to create a histogram of the mixed channel. The amount of stretching for the mixed channel is proportionally applied to the individual color channels. Therefore, the automatic manipulation based on all channels should not result in a color cast. Individual Adjustment for Each Channel Each histogram is stretched individually not considering the other bands. Therefore, this automatic manipulation usually will also change the color which in most cases is not wanted.
Images
Note that the adjustments are only affecting selected images. However, different options are available to complete the necessary histogram manipulations. Common Adjustment Based On All Images Analyzes the histograms of all images no matter, if selected or not. This option might be used to adapt few selected images to the rest of the block. Common Adjustment Based On Selected Images Analyzes the histograms of the selected images only. This option might be useful if by nature the selected images have a different color or intensity compared to the rest of the block, e.g. water images. Individual Adjustment for Each Channel
Page 34
Performs a histogram manipulation for each image separately without considering all other images. This option might produce a block of images that are looking very inhomogeneous.
Magnitude of Correction:
According to the settings for all images or all selected images the software computes the min/max values of the image histogram. The auto-adjustment modifies the histograms of the images by considering the magnitude of correction.
Figure 18:
Original histogram of one or all selected images
Figure 19:
Correction with 100% magnitude
In this case, the histogram is stretched using the full extents of the color range (0256 respectively 0-65535).
Figure 20:
Correction with 90% magnitude
In this case, the difference between the min/max values of the histogram to the maximum extents is computed and the histogram is stretched to 90% of the missing data range. Note: The gradation and intensity correction overwrites enhancements made with the contrast/intensity, saturation and color editor. As a general rule: If you intend to work with all functions of the Radiometrix Editor, always work through the tabs from left to right.
Cutoff Percentage
This option allows to cutoff a certain percentage of the dark and bright pixels of the histograms before they are stretched. The option helps to automatically enhance
Page 35
the images much better as there might be noise on the dark and bright side of the histogram that might badly influence the histogram and therewith the images. Note: Do not cutoff too much as you might loose image content.
Interactive Gradation Curve and Intensity correction

This function allows modifying the gradation curves or intensity of only selected images. The changes can be made on individual channels or on the mix channel. Gradation curves are the most flexible tool for image enhancements. The advantage is, that the darkest and brightest image values are not changed (compared to stretching a histogram too much), so that no texture is lost.
--> X The x-axis of the diagram represents the input of the pixels. The y-axis represents the output values. The dark values (0) are on the left. The linear line represents identical input and output values. To change the color balance/intensity, first select the images you want to modify (See Modifying Images below). Then select the channel you want to modify and activate the Auto Preview button to see the changes in the main window immediately. To change values, click on the curve with the left mouse button. Keep the mouse button pressed and release the button on the desired position. Doing this allows to add as many points as necessary. Each point represents a vertex of a polygon. The resulting modified curve is a B-spline.
To modify vertex points, click on the point with the left mouse button and drag the point by keeping the mouse button pressed. To delete vertex points, click with the right mouse button on the point to delete. The Apply button saves all changes in the corresponding rdx files of the images and the curve is then changed back to a straight line. Note: The gradation and intensity correction overwrites enhancements made with the contrast/intensity, saturation and color editor. As a general rule: If you intend to work with all functions of the Radiometrix Editor, always work through the tabs from left to right.
Page 36
Examples:
Figure 21: Situation before working with the gradation correction
Figure 22: Correcting all channels images are to bright
Figure 23: Correcting all channels images are now darker
Page 37
Figure 24: Correcting all channels images are now darker and have more contrast
Figure 25: Correcting upper row of images and blue channel only too blue images are better adjusted to the other images
Modifying contrast/brightness/color/saturation of images

These editors are more intuitive to use than editing gradation curves however, be aware that gradation curves are more flexible to use. To modify images: Select images to be changed. The corresponding dots in the Radiometrix Editor will be highlighted and the center coordinates of the selected images are displayed with Selection Location. Select Modify Click with the left mouse button on a position in the Radiometrix window. This position will be the new position of the group of selected images and the selected images are respectively changed.
Page 38
New location
Figure 26: Situation before and after modification with pointer to new position Relative Mode shifts the average value for the selected images while preserving the relative relationship (and differences) between them. Absolute Mode moves each selected image to the same position. This will cause all the selected images to have the same average value. This can be desirable in giving the group a consistent look; however, it can also shift images away from true relative colors. For example, moving the color space of an image containing a brown field to the same colors of an image of a grove of green trees can cause the brown field to look green.
new position
Original Position of selection

new position
relative modification
new position
50 % relative/absolute
absolute modification
Now go to the various Tabs with the mouse, and reposition the desired selected images by selecting the new location with a left mouse button click. The selected images will move to the new location and if the main window is displaying the input orthos, the color or contrast/intensity or saturation of the selected images will be changed. Continue with changing the images until you are satisfied with the enhancements.
4.19.2 Save/Reject changes of the Radiometrix Editor

By pressing the Ok button all changes are saved for further use and the Radiometrix Editor will be terminated. By pressing the Cancel button all changes are rejected and the Radiometrix Editor will be terminated. By pressing the Undo Changes button any changes made to images since starting the Radiometrix Editor will be undone.
Page 39
By pressing the Reset to Source button all changes are reset and the original color/intensity/contrast values are recomputed from the source images.
4.19.3 Radiometrix Editor: Background information

When starting the Radiometrix Editor for the first time, OrthoVista generates a rdx file for each image, on the directory where the images are stored. This makes it necessary that the protection setting for this directory is write enabled. When the Radiometrix Editor is started afterwards, it reads all changes stored in the rdx files. All changes done with the Radiometrix Editor will be taken into account when processing the images. Note they are even used if you have defined none for Single image and Group image adjustment under Processing Options.
4.20 Color Picker

The color picker allows checking the color of a certain pixel or the mean color of a selected area.
Figure 27: Color Picker Window To pick a color of a pixel or a mean color of an area the Pick Point/ Select Area button must be activated first. Then you have to either select a single pixel by clicking with the left mouse button or drag a rectangle. The color picker displays the RGB color values of the pixel or the mean value of the selected area. If several images are overlapping each other at the selected position, the software displays the mean value of all overlapping images. Once such a color is available, it can be assigned to the parameter settings Display Background Color or Output Background Color or Reflections Removal Base Color with the Use as button. If the images are changed with the Radiometrix tool the color picker will immediately change the color according to the changes triggered by the Radiometrix Tool. Multiple Color Picker instances can be active at a time. Each Color Picker indicates its position by a small circle (for points) or a rectangle (for areas).
4.21 Status Dialog

The Status Dialog informs about the current status of the processing but also contains information about start and termination of a process. In addition it
Page 40
displays warnings, errors and messages. The status dialog is automatically opened as soon as OrthoVista writes information into the status dialog.
Page 41
Processing Options
Once your imagery is loaded and the processing area is defined, you can process imagery using OrthoVista. Almost all image-processing options are controlled in the Processing Options dialog. To begin processing: Select Begin Processing
Figure 28: Processing Options dialog The Processing Options dialog maintains and displays the settings from your last processing run, so you only need to decide which settings to change from your previous session. However, the Output Directory location is not saved. The output directory has to be defined with each start of the Processing Option Dialog. To start processing, click the Close And Process button. OrthoVista processes the images defined by Tile definition or Area definition. A status window shows the status of the processing. When the processing is complete, a small dialog window notifies you that your output images are available. The processing options you specify are automatically saved for the next processing run.
Page 42
The Close button closes the window without processing but saves the current settings. The Cancel button closes the window without processing and without saving the current settings.
5.1 Output options

5.1.1 Specifying the Output Directory
The Output Directory is the directory into which the adjusted imagery and/or new mosaics will be written. To specify the output directory either: Click in the Output Directory box and type the appropriate path for the directory. Click the Browse button and use the Select Directory dialog to browse for the desired output directory. Remember to specify an Output Directory location different from the input image directory location.
5.1.2 Meta Data Directory

OrthoVista generates the following meta data files. .rgn (region) files contain information about valid image areas. Valid image areas are those areas that do not include background color information .tlt (global tilting) files contain information generated by the global tilting adjustment .bal or .spb (single image balancing) files contain information generated by the single image adjustment cld and fda files contain the seam lines and blending information for the Feature Detection mode cld and agd files contain the seam lines and blending information for the Adaptive Feathering mode autoseam.trn files contains additional blending information for Adaptive Feathering Other interim files are generated during the processing and after a normal stop of OrthoVista automatically removed. These files are interim files that are not important for later processing. The definition of the Meta Data directory is no longer optional. If not defined, OrthoVista computes the above-mentioned data and stores it on the same directory, as the input files are located. By default OrthoVista defines a meta data directory called meta which is located on the output directory. The purpose of the Meta Data directory is: Speed up processing. Data is computed then stored and used later when needed again. For instance, the region (rgn) information is needed in OrthoVista and Seam Editor. If you run OrthoVista before using Seam Editor and store the meta data, Seam Editor has not to re-compute the information and therefore the project setup is much faster. Transferring balancing and tilting information from one sub-block to another. For instance if you have to process a large block in two separate parts it is
Page 43
helpful to process the first part and store the meta data. Then to process the second part later on with overlapping images to the first part and use the same meta data directory. Meta data, which is already available, is not re-computed but used. With the help of this you transfer then tilting and balancing information from part one to part two. The result is that the radiometric difference between the two parts would be lower as compared to running the two parts independent from each other. Note: Do not use the same Meta Data directory if you just want to re-run OrthoVista with different parameters, as OrthoVista does not re-compute the information if it exists, resulting in no difference between the different runs.
5.1.3 Output Image format

OrthoVista supports a variety of image formats. TIFF and BIP/BIL/BSQ are standard, but other formats may be supported through the use of plugins. To find out the specific image formats supported by your installation of OrthoVista, select About Plugins from the Info menu and select Image Support Plugins. To specify the image file format for the output images, select the desired format from the Image Format drop-down box. Each image format also has a variety of options that can be accessed by clicking the Options button next to the Image Format drop-down box. Note: When changing the image format, the extension for the output filenames will be changed according to the format to reduce confusion and incompatibility with other image programs. Modifying software configuration options may overwrite the specific extension for each file type. For the TIFF format, additional options may be specified from the TIFF Options dialog.
Figure 29: TIFF Options dialogs. Options for TIFF image format:
Page 44
Layout: Supports Tiled TIFF and Scanline TIFF. Sample Format: Supports a variety of Tiff formats. Overview: When checked, OrthoVista includes one down-sampled image overview or a full set of down-sampled images as supported by the TIFF format. The Overviews are stored either in a separate file or within the output file itself. The separate files will be located on the same directory as the output files. The overviews, especially a full set of overviews make the display faster for software that supports this TIFF capability.
For the BIP/BIL/BSQ format, additional options may be specified from the Options dialog.
Figure 30: BIP/BIL/BSQ Options dialog. Layout: Supports band interleaved by pixel or line or sequential.
5.1.4 Output Report format

To specify the georeference file format for the output reports, select the desired format from the Report Format drop-down box. Some report formats have a variety of options that can be accessed by clicking the Options button next to the Report Format drop-down box. GeoTIFF Note: OrthoVista handles GeoTIFF files that contain the image and header data in the same file and supports the following tags in GeoTIFF format: ModelTiepointTag ModelPixelScaleTag ModelTransformationTag Other tags that describe projections as well as extensions that are not supported will be carried through the process and written in the output files without alteration. More information concerning GeoTIFF can be obtained from OrthoVistas Info menu. The fractional precision and notation of coordinates stored in the TiffWorld file can be changed with optional parameters.
Page 45
Figure 31: TiffWorld file format options The notation style options are Fractional which generates coordinates with a notation like 12.123456 Exponential which generates coordinates with a notation like 0.12123456e4 Automatic generates dependent on the coordinates either the fractional or exponential format
5.1.5 Save Background Information for Output Images

If this option is activated, OrthoVista automatically generates the region files (rgn files) for the output images. This greatly reduces the time for activating the images in the Seam Editor, if and only if you load in the Seam Editor your output images. What you only do when you use the Save Adjusted Images option and you load the adjusted images in the Seam Editor or when you do the mosaic file editing in the Seam Editor. In most cases you dont need this function.
5.2 Adjustments options

5.2.1 Specifying radiometric adjustments for single images
The following radiometric adjustment methods are available for processing individual images and are described in detail in the following paragraphs: None Hot Spot Removal Intensity Dodging Hint: From experience with many data, sets we can say that Hot Spot removal and Intensity Dodging should be only applied when your images show lens vignetting effects or color variances. If either Hot Spot removal or Intensity Dodging has to be applied, the Hot Spot Removal function works better for color images and the Intensity Dodging function better for black and white images.
NONE
If you select None, single image processing will not be performed on individual images. This is the default and should be selected whenever possible.
Hot Spot removal

Applies correction to compensate for effects such as illumination hot spots and lens vignetting. See also chapter 6.5. Corrections are applied to each input image
Page 46
individually. For hot spot removal, additional options may be specified from the Hot Spot Removal dialog.
Figure 32: Hot Spot Removal dialog. Hot spot removal Options: Images with large Background Areas Should be activated if the images have many pixels with background color. If this option is not activated but the images have large background areas, the relevant images will be skipped for Hot Spot Removal Sampling Grid Size. The number in this box defines the number of samples taken in each of two directions from each of the input images to derive the optimal balancing parameters for each image. The grid size value affects the quality of the balancing correction in the output image. Larger values produce better quality but require more processing time. The data range is 40-300 if images with large background areas is deactivated and 50-2000 if images with large background areas is activated. Note: At least 1500 valid samples are required for the Hot Spot Removal Example: An image is covered with about 20% valid image pixels (80% background color). 1500 samples need to cover 20% of the image 1500 samples = 20% 7500 samples = 100% Sqrt(7500) = ~87 => 100 To make sure enough samples can be derived, a sampling grid size of 100 should be defined. Method. The method defines the algorithm used to perform the radiometric
Page 47
corrections. Often, one method is better than the other for certain types of projects. Additive is generally much faster than Multiplicative (see the Additive vs. Multiplicative section in Chapter 6 Advanced Techniques). Maintain Intensity. The Maintain Average Input Intensity checkbox defines whether the software will ensure that the average intensity of the output images after radiometric correction will match the average intensity of the input images. Average Intensity on best preserves input image intensity. Apply Color Correction (For Color Images Only). When enabled, individual image processing includes correction for systematic color trends within individual images. If disabled, color remains unaffected and only the intensity is modified during processing. Excluded Area. OrthoVista does not perform radiometric balancing to the area within a polygon that is defined by an ArcShape file or DXF file specified in the Definition File box. If imported by DXF files the polygons must be stored in a layer with the capital later named EXCLUSION. This capability is useful for dealing with processing problems related to water bodies (see the Using user-defined vectors section of Chapter 6 Advanced Techniques).
Intensity Dodging
Performs intensity modification to remove complex intensity variation within images. Correction are applied to each input image individually. See also chapter 6.6. For Intensity Dodging, additional options may be specified from the Dodging options dialog.
Figure 33: Dodging options dialog Intensity dodging options: Grid size and Sample Size. Together the two "Sampling Grid" options work as follows. Assume only one dimension (e.g. across a row) for illustration - but the algorithm works in two dimensions. For illustration, assume a (very) small image with 36 by 36 pixels. If GridSize=4, this image will be sampled at 4 equally spaced locations in each
Page 48
direction. At each location, many of pixels are sampled. SampleSize determines the number of pixels sampled at each location. E.g. if SampleSize=3, then three adjacent input image pixels will be used in the computation. Of course the sampling is done in each of 2 dimensions, so that a SampleSize=3 actually includes 9 pixels in the computation at each of the 16 (4x4) locations. So for this example, there are a total of 1296 pixels of which 144 pixels (3x3 x 4x4) are used in the computation. This example is illustrated graphically in the following: The "+" indicates the location on which each sample is centered (remember this will actually be done in each dimension - so that for a 36x36 pixel image, a grid of 4x4 locations will be selected. ----+--------+--------+--------+---At each location, a sample of contiguous pixels is taken. Below, the "X" indicates pixels used in the computation. In two dimensions, a patch (e.g. 3x3 for this example) is sampled. ---XXX------XXX------XXX------XXX--If the SampleSize parameter is 0 or negative (e.g. <=0), then the samples expand until they touch the adjacent samples. Therefore, in such a case (the default), ALL pixels of the input image are sampled. ----+--------+--------+--------+---XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX Note that the location points, still determine the amount of detail used in the computation ( e.g. order of the mathematic model being fit). Sample Size defaults to value "-1" which means to use *ALL* pixels of the input image (compatible with previous behavior). This can be set to a small value to provide faster image dodging -- e.g. 200 or perhaps 500 should work fine in many cases (unless there are unusually large bright/dark areas in the image) and will run faster. But to get best results we suggest using the parameter 1. Method. The method defines the algorithm used to perform the radiometric corrections. Often, one method is better than the other for certain types of projects (See the Intensity Dodging section in Chapter 6 Advanced Topics). Maintain Intensity. The Maintain Average Input Intensity checkbox defines whether the software will ensure that the average intensity of the output images after radiometric correction will match the average intensity of the input images. Apply Color Correction (For Color Images Only). This option is not available for this method.
5.2.2 Per-Image Selection

If Hot Spot removal or Intensity Dodging is applied, it is possible to define on a per image bases in which image the functions shall be applied. By default all images are selected.
Page 49
Figure 34: Individual Image selection box for applying single image adjustment
5.2.3 Image group adjustment

Image group adjustment allows applying two different functions. The Global Tilting Adjustment function compares images radiometrically in the overlapping areas and computes with the comparison result radiometric adjustment parameters for the images. The Reflections Removal function tries to detect and eliminate e.g. sun reflections on water areas. Note: Both functions cannot be applied during the same OrthoVista processing run since they would interfere with each other. We suggest that you first process with the Reflections Removal option and use then the resulting images for the global tilting adjustment. 5.2.3.1 Global Tilting Adjustment The following radiometric adjustment methods are available to compute corrections to compensate for intensity/color/contrast variation between adjacent images and are described in detail in the following paragraphs: Global Tilting Adjustment None If you select None, radiometric image processing will not be performed between adjacent/overlapping images. Global Tilting Adjustment performs radiometric adjustment to compensate for intensity/color differences between adjacent/overlapping images. For the global tilting adjustment, additional options may be specified from the Tilt Options dialog.
Page 50
Figure 35: Global Tilting Adjustment dialog. Options for global tilting adjustment: Adjustment Iterations. The global tilting adjustment is an iterative algorithm. This number determines the number of iterations that are performed. A value of 2 or 3 is usually sufficient. Larger values increase processing time but can be helpful in solving light falloff effects. Grid Size Global Tilting is using a grid cells to compare overlapping parts of the images for computing the adjustment values. By default a grid size of 80 by 80 cells is used. The Slider allows to alternate the default (slider in the center). When set to Coarse 20 by 20 grid cells are used. When set to Fine 140 by 140 cells are used. For images with large radiometric differences we suggest to increase the default value up to fine. Images with low radiometric differences or with a large overlap we suggest to lower the default up to coarse. The larger the value the higher the processing time of the global tilting process. "Hold extreme corners at input values": If activated, the image corners located at the extreme corners of the mosaic are not radiometrically changed. Default is Do not hold extreme corners. This parameter should only be activated in cases where the global tilting adjustment produces extreme corners that are too bright or too dark. Review the following Figure. Input for the example are 4 single images upper left was dark upper right was white lower right was gray lower left was dark at the lower left edge and gray at the upper right edge.
Page 51
Figure 36: Hold extreme corners
Do not hold extreme corners
Adjust Color If this option is activated then the software performs an intensity/color adjustment.
Figure 37: Before (left) and after (right) intensity/color adjustment Adjust Contrast When activated the software performs a contrast adjustment. See chapter 5.2.3.2 for a detailed explanation of the function and its parameters.
Figure 38: Before (left) and after (right) intensity/color/contrast adjustment Maximum Gray and Color Difference The global tilting adjustment is done by comparing images in the overlapping areas. The difference between the images is used to compute correction values. Global tilting might lead to unsatisfactory results in cases where the image content in one image which does not occur in the overlapping image or at least not on the same position. Good examples of this are clouds and sun
Page 52
reflections on water. To avoid this problem, adjustment options provide a method to define percentage levels of gray and color differences. This value determines if a gray or color difference at a certain position shall be used or excluded for the tilting adjustment. Figure 35: shows the default values. Note: Define these values only, if you do have images containing extended clouds or sun reflections on extended water surface. These two parameters are newly introduced with version 4 of OrthoVista. If you would like to have the same processing behavior as with former versions, you have to define 100 percent. 5.2.3.2 Contrast adjustment options The Contrast Adjustment allows reducing or eliminating brightness/contrast differences between images from north to south and east to west. Such differences are mainly caused by shadows. Such brightness/contrast differences in the images can be already greatly reduced by generating orthophotos containing image information from the center of the aerial images and having enough image information so that orthophotos overlap but do not overlap too much.
Figure 39: Orthophotos with full extent and reduced extent The above figure shows on the left a data set with orthophotos having the full extent of the aerial images. The big brightness/contrast differences can be clearly seen in east-west and north-south direction. On the right it shows a data set processed with OrthoMaster containing only information of the center of the aerial images but having an extent large enough so that the orthophotos overlap by about 15%. In this data set the differences in eastwest direction are already eliminated. Still differences in north south are available but compared to the left example the differences are smaller and therefore OrthoVista will not have to make such adjustments. Note: Contrast adjustment may result in over-saturated images. Please use with care. Note: If you do not have east-west and north-south contrast differences, then dont enable the contrast adjustment as it takes more processing time to do this kind of adjustment.
Page 53
Figure 40: Global Tilting adjustment options with Contrast adjustment
Maximum Contrast Difference The contrast adjustment is done by comparing images in the overlapping areas. The difference between the images is used to compute correction values. The contrast adjustment might lead to unsatisfactory results in cases where the image content of the overlapping images is too different (e.g. one image shows a cloud where others show the ground). If this option is activated and the contrast difference is above the limit then the observations at a certain position are not considered for the adjustment. Contrast Ratio Contrast Ratio defines a scaling factor for the correction value computed with the adjustment. Because of speed issues the correction values are computed with the overview levels. As in overviews always the contrast is reduced compared to the original image level, the correction values might be too low. The Contrast Ratio 100% means that the correction values are used as computed. A value of e.g. 120% defines that the computed contrast values are scaled with a factor of 1.2. The valid range is 50% to 200%. Contrast Target The function computes for all the images the contrast values. The Contrast Target defines now the target value for the new contrast. Minimum means in principle that the image with the lowest contrast defines what shall be achieved. Whereas maximum means that the image with the highest contrast value defines what shall be achieved. This parameter therefore allows to define if the resulting files shall have a more strong or soft contrast.
Page 54
Figure 41: Example of contrast adjustment with the settings above 5.2.3.3 Reflections Removal The reflections removal function checks for sun reflections and tries to eliminate them. The area which is checked and modified for sun reflection must be defined by a DXF file. When sun reflection is found, the area is smoothed to filter out the reflection. The smoothing can be influenced by changing the smoothing degree from low to high. In addition to the smoothing, the area can be filled with a certain color which represents the mean color of the water area. Along the given water area the software is blending the area with the surrounding area to get a smooth transition between water and non-water areas. The blending distance to the left and right is given in pixels.
Figure 42: Reflections Removal Options Texture Smoothing First of all, the Reflections Removal function eliminates only the reflections, but structures such as waves and underwater elements within the images are kept. These structures can be smoothed with the texture smoothing filter.
Page 55
The above figures show from left to write an original orthophoto section with sun reflection on the water. The next image shows then the result of the Reflections Removal function with low filtering and further on with high filtering. Intensity Only The following figure shows again the section processed with intensity only.
Color artefact
In some cases, this method may cause color artifacts (we used here CIR images which contained a high red color cast). To avoid such color artifacts, you should use a base color to process the reflections removal. Use Base Color
The above image was processed with a base color and a low texture smoothing. The sun reflection is completely removed, with no added color artifacts as they can occur with the Intensity Only option. The base color can be easily defined with the Color Picker (See chapter 4.20).
Page 56
Feather Distance
The Feather Distance defines the blending area in pixels on each side of the to be defined Reflection Area. Reflection Area Definition File It is necessary to define the Reflection Areas with closed polygons. Several areas can be defined. If a defined area is within another area, the inner area defines an island which is then treated as an exclusion area for the Reflection Removal. Currently the Reflection Areas must be imported via a DXF file or ArcShape file. The polygons in the DXF file must be stored in a layer with the name REGION. Allow Complex Polygon Definitions When the option Allow Complex Polygon Definitions is enabled then the import routine analyses all the polygons. To skip the analysis, the polygon definition needs to fulfill two requirements: (1) The polygon must not be self-intersecting, i.e. there must not be any line segments that intersect each other. In case partly overlapping areas are found, they are merged together to one common polygon if the Allow Complex Polygon Definition is switched on.
Figure 43: Two polygons overlapping
Result Complex Polygon activated
(2) The vertex order of the digitized polygons is important. Water Area Land Area/Islands
Figure 44: Recommendation: Water area digitized counter-clockwise

Page 57
If Complex Polygon is not activated the direction how polygons are digitized is important. In this case digitize the outer water areas counter-clockwise and all islands clock wise. If a polygon with the type left exclusive is used, then the water area may also be digitized clockwise and the islands have to be digitized counterclockwise. If the polygon definition does not satisfy both requirements, the flag Allow Complex Polygon Definition must be set. Note: In case of Allow Complex Polygon Definition - for large Reflection Area Definition files the analyzing can take a considerable time (several hours). In case partly overlapping areas are found, they are merged together to one common polygon. So enable this option only if you would like to combine polygons to one and you dont like to consider the digitization direction.
5.2.4 Mosaic adjustment

The Mosaic Method drop-down box allows you to specify the algorithm used to fuse individual images into a mosaic: Plain Mosaic Sheet Centered Adaptive Feathering Seam Applicator
Plain Mosaic
Generates mosaics by arbitrarily placing one image on top of the other. It is the fastest mosaicking method; however, geometric seams are restricted to image edges and no feathering is performed. This method should be used if you already have fitting mosaics and you want to combine them to larger mosaics.
Sheet Centered
Generates an image in which all data for a single output tile are drawn from a single input image. This means that within an output tile there is no mixing of data from multiple input images. For this to be effective, orthos should be produced on a sheet-centered basis, and the completely contained sheets should be within one input orthophoto. This method allows you to color balance and radiometrically adjust images and then to crop each one to an appropriate sheet extent. The result is a set of sheets, which are produced with pixels from only one image. However, the images will have been radiometrically adjusted so that the resulting sheets will match when reassembled. This method should be used if you want to cut out images from existing images and you want to be sure that data be taken from a certain image.
Adaptive Feathering
Merges individual input images into a seamless mosaic. This method performs digital adaptive feathering that replaces conventional manual procedures. The method automatically computes a blending function, which determines how to combine the individual input images into the output mosaic. A blending sharpness parameter controls the steepness of the blending function, which in turn controls
Page 58
the default width of the blending function. In areas of complex relief displacement (e.g. buildings), the blending function is automatically made steeper, so that the image-to-image transition occurs more quickly in these regions. Additional options may be specified from the Adaptive Feathering Options dialog.
Figure 45: Adaptive Feathering Options dialog Options for adaptive feathering: Support True-Ortho Mosaic Enable this option if you have to process orthophotos that do have small background pixel areas within the images (e.g. True-Orthophotos). If not enabled, the software might not detect these areas and might then not fill them up or blend them with other overlapping images. Area Type The area type gives parameter setting suggestions for different terrain types. User defined allows modifying the parameters as you like. If you select a certain terrain type the parameters are fixed to INPHOs suggestions. Alpha Grid Size. Determines the amount of detail considered during the adaptive feathering computations. Larger values produce a more detailed and complex transition boundary between overlapping images. Note that the processing time is strongly dependent on this value. Hint: Use the default settings Urban, Mixed, Rural dependent on the terrain type. As for images which are much more long than wide, like the ADS40 images, the standard settings are not optimal. For these type of files 3 default settings called ADS40 urban, ADS40 mixed and ADS40 rural are selectable. Sharpness. Controls the sharpness with which one image blends into another. Larger values produce a quicker transition. For example a value of 0 (not typically used) would produce an equal weighting of images throughout the overlap region. A value of approximately 2.5 produces a very smooth transition (similar feathering across an approximately 1/5 of the overlap region). A larger value such as 20 produces a very sharp transition within approximately a pixel width. Hint: We suggest using a Blending Sharpness of 3 with small-scale orthos and a Blending Sharpness of 5 to 7 in urban areas or large-scale orthos. For forest areas use a Blending Sharpness value of 4 or 5. The following examples show an urban area with an image scale 1:4000. It is obvious that a grid width of 80 is too low since cars are not detected and the seam line cuts the cars. The blending sharpness setting of 2.5 is too wide for this area type. Cars would be blended which would lead to ghosting effects.
Page 59
Figure 46: Left image: Right image:
Grid Size 200 and Blending Sharpness 2.5 Grid Size 200 and Blending Sharpness 7
Figure 47: Processed with: Grid Size 80 and Blending Sharpness 7
Feature Detection
The Feature Detection function is a second method allowing automatically derived seam lines. Feature Detection is better adapted to urban areas than the Adaptive Feathering function, and it can also be used in different areas such as mixed and rural. The Feature Detection function is in general slightly faster than Adaptive Feathering with seam lines avoiding more features, like buildings, especially in urban areas. Parameters of the Feature Detection function:
Page 60
Figure 48: Feature Detection parameters Area Type The area type gives parameter setting suggestions for different terrain types. User defined allows modifying the parameters as you like. If you select a certain terrain type the parameters are fixed to INPHOs suggestions. Feature Size Feature Detection works by subdividing an image into small cells and comparing the radiometric content of these cells. Cells that fit together are used to define the seam line. The parameter Feature Size defines the size of a quadratic cell in pixels. The cells should be small enough to contain details of features being not mixed up with other features. On the other side the cell should not be too small to avoid long processing times. Feature Granularity Defines with which image resolution the processing is done. The following three granularities are offered. Fine Processing on image resolution 1:1 Normal Processing on image resolution 1:2 Coarse Processing on image resolution 1:4 Use Maximum Values For each cell radiometric parameters are computed. Either the mean or maximum values are used to compare cells with each other. Minimum Coverage Determines if a cell is considered being part of the overlap or not. A cell is part of the overlap if more than Minimum Coverage in percent of the pixels in this cell are valid pixels in overlapping images. Blending Threshold Specifies if a cell is blended or not. If the computed value is lower than the
Page 61
threshold, the cell is blended between overlapping images. Smaller values lead to sharper blending, while higher values may introduce ghosting effects. Use Edge Detection Specifies if an edge detection method is used in addition to the feature detection method to define the seam lines. Edge Detection Weight Determines the weight of the edge detection differences compared to the radiometric differences. A value of 1 defines an even weighting of edge detection and radiometric difference. Add Border Distance Specifies if an adaptive border distance should be added or not. If activated an adaptive border distance is added to keep the Seamline in the middle area of the overlap. It can be deactivated if the images have less overlap or if mosaic hints are used.
Seam Applicator
Generates seamless mosaics with the help of seam polygons defined with the Seam Editor The complete process of mosaicking with manually defined seams consists of two main steps: 1) define a polygon partition for the mosaic with OrthoVista SE program; and 2) apply that polygon definition during OrthoVista processing via the Seam Applicator. The polygon data are transferred between the two steps via the MetaData Directory. Here it is assumed that a polygon partition has been defined, and that the partition data have been stored in a MetaData Directory of your choosing (See OrthoVista SE documentation). In order to run the Seam Applicator it is necessary to define the MetaData Directory on which the Seam Editor has stored the seam polygons. This is accomplished by specifying a directory in the OrthoVista processing options dialog (See section: Meta Data Directory). Complete the other fields in the processing options dialog, as you would normally do - except, ensure that you select "Seam Applicator" as the mosaicking option. Additional Seam Applicator options may be specified from the Seam Applicator Options dialog. When Seam Applicator has been selected, this options dialog can be accessed from the "Options" button next to the mosaicking method selection box.
Page 62
Figure 49: Seam Applicator Option Window Feather Size The feather size is given in pixels and defines the blending width with which one image blends into another. With the feather mode, you define how the blending is performed. Two options are available: Linear Linear blending between the overlapping images
Inverse Distance Blending with a functional curve dependent on the blending width and radiometric differences between the images
Blend Output Area Borders By default the blending is only done along the seam lines, which are interactively defined in the Seam Editor. If you activate this mode then OrthoVista will blend also along the orthophoto border, which might lead to unwanted effects. Tip: If you know beforehand that you will have to quality check and likely edit the automatically detected seam polygons generated by the Adaptive Feathering run, then you can save time by using the Adaptive Feathering and run the process to generate the seam polygons but not the mosaics themselves. You do not absolutely need the mosaics for the seam editing. Note: When using rotated images in Seam Applicator, the OrthoVista display has to be set to the heading corresponding to the angle of the images. The output area then has to be defined manually, as simple tile definition and tile specification do not support rotated tiles.
5.2.5 Mosaic Hints

Mosaic hints mark exclusion areas for seam lines. The automatic seam line finding functions (Adaptive Feathering, Feature detection) will then avoid these areas. Import of closed polygons in DXF and SHP format are supported. Note: If mosaic hints are used with Adaptive Feathering, the alpha grid size needs to be at least urban (200) or higher. Note: If an exclusion area covers the complete overlap, the seam line may go straight.
Page 63
5.3 Output Selection

5.3.1 Saving adjusted images
Activate the Save Adjusted Images checkbox to output radiometrically corrected versions of individual input images which can be viewed and used independently of the mosaic. This checkbox does not effect processing of the mosaic. The Options dialog defines further on how the mosaics are saved.
5.3.2 Rotate adjusted images before saving

OrthoVista provides the Rotate Adjusted Images Before Saving checkbox for handling rotated imagery. When checked, the images are resampled according to the angle setting in the canvas orientation window and then saved on disk. Note: If your input images are not rotated to north and you have to process the images, OrthoVista has to resample the images several times during processing which increases the processing time by several factors. Therefore we suggest to first rotate them to 0 degree, by setting the single and group image adjustment to None, switching off the mosaicking, switching on Save Adjusted images, defining a Canvas orientation of 0 degree and then processing the images. This also takes some time, and you need more disk space, but subsequent processing will be much faster.
5.3.3 Generate Seam Data (*.cld Files)

If activated OrthoVista generates and saves the Seam Data as cut line definition files (*.cld) to the meta directory. The export selection offers the choice between the generation of the cld files only and both cld and DXF files. Note: This option must be activated if later on you would like to edit the seam polygons with the Seam Editor. Saving seam data increases the overall processing time by about a factor of 10 to 15%.
5.3.4 Save Vector Seams

OrthoVista allows saving the automatically generated seams (e.g. by Adaptive Feathering) in a DXF file. The seams are exported as 2D Polyline into the Layer SIMPLIFIED. In addition to this a boundary around the whole area is exported. This boundary is stored in the layer BOUNDARY. The purpose of the seam export is to document the results and to overlay the DXF data on top of the mosaics in order to check the result.
5.3.5 Save Vector Seams for each image

If activated OrthoVista saves the seam lines in individual DXF files for each image. These DXF files could be imported in the Seam Editor e.g. for images flown at a later time to use exactly the same seam lines as with the first mosaicking process.
Page 64
Note: this works only if the images have the same size, position and image border pixels as the originally used images.
5.3.6 Seam simplification tolerance

OrthoVista internally treats seams pixel-wise. If this kind of data would be exported to DXF, the file size would be very large. Therefore the Save Vector Seams function simplifies the pixel information and changes it into polyline. The Seam Simplification Tolerance defines the maximum displacement of the polygon against the individual pixel positions. Default is a simplification factor of 3.
5.3.7 Save Mosaic Output

If enabled then OrthoVista saves according to the settings for Mosaic Adjustment mosaics, also called tiles. The Options dialog defines further on how the mosaics are saved.
5.3.8 Options for Saving Adjusted Images and Mosaic Output
Figure 50: Output options The output options define the File Naming Format, the Output Ratio, number of channels to be written and allows to assign channels. Several definitions can be made and allow to Save at the same time different file formats in different channel combinations. 5.3.8.1 Internal Name of Output Several different data formats can be defined. For each definition, OrthoVista will later on save on different directories adjusted images and/or mosaics. New definitions are made by pressing the Add button. The Remove button allows removing already defined names. The Default setting cant be removed. 5.3.8.2 Directory The Directory display shows on which directory the files are stored. For each Internal Name definition a subdirectory below the output directory will be created. Only the default settings are saved on the output directory.
Page 65
5.3.8.3 File Name Format %s is a placeholder in case of Save Adjusted Images options for the input file name and in case of Save Mosaic for the tileID. The file naming format can be modified by adding alphanumerical characters in front or behind the %s placeholder. Never remove %s, as then your output file names are no more unique and OrthoVista will then not generate all files but overwrites them several times. Example: Input file name 1 = orthophoto1.tif Input file name 2 = orthophoto2.tif %s is then the placeholder for orthophoto1 and orthophoto2. In case of output format is set to %s Save Adjusted Images would save 2 orthophotos with the names orthophoto1.tif and orthophoto2.tif. In case of output format is set to abc_%s_north Save Adjusted Images would save 2 orthophotos with the names abc_orthophoto1_north.tif and abc_orthophoto2_north.tif. 5.3.8.4 Ratio You can use OrthoVista to generate down-sampled output images. To do this, enter an integer number larger than 1 to generate output images at a scale smaller than that of the input images. For example, if you type 2, 2 by 2 pixels are taken to resample a new pixel, which means that the width and height are divided by 2 and the image size is reduced by factor 4 (2x2 pixels). If you activate the Average pixels during down-sample toggle, averaging input pixel neighborhoods will produce the output pixels. This results in a smootherappearing result and should be usually activated. 5.3.8.5 Number of channels and RGB component setting The parameter Number of Channels defines how many channels shall be saved. It allows e.g. to convert a 4 channel image to a 3 channel image. Output RGB components defines if the output file shall have the information that the output file contains RGB channels or not. 5.3.8.6 Channel assignment The channel assignment table allows defining which input channel shall be mapped to which output channel. 5.3.8.7 Example for 4 Channel RGB and Infrared image In the following it is assumed that the input images are multi-channel images containing 4 channels. The first three channels are RGB and the fourth channel contains Infrared information. The output shall provide two different image types containing just the RGB information and in another file containing the Color Infrared images. When the Options dialog is called the first time it comes up with the default settings.
Page 66
Figure 51: Default settings of 4 channel RGB-IR images In a first step the default settings are now modified, so that by default only RGB images are processed.
Figure 52: Modified Default settings to save RGB images In a next step a new definition is made for an additional color infrared image (CIR) output.
Figure 53: Settings for a CIR output Note: For the RGB output no new assignment was done, but the default setting was changed. This is necessary as the default setting cant be removed and will be always processed. If a further RGB definition would
Page 67
have been defined instead of modifying the default setting, OrthoVista would have exported 4 types of images - 4 channel images, RGB and CIR.
Page 68
Advanced Information
OrthoVista supports multi-channel images. If the multi-channel images contain RGB channels then it is important to define the RGB channels with the Image Commander in a first step. The RGB channels are treated then as an RGB image whereas all other channels are treated as single bands. Should you have to process RGB images which are stored in 3 band images like ERDAS Imagine is doing this, you should also use the Image Commander to define the RGB channels. Processing of single bands of multi-channel images. The effect is the same as if: 1) original image bands were separated into separate grayscale images; 2) each of the separated grayscale images was adjusted in a single channel mosaic; 3) the single band mosaics were recomposed into a multi-channel mosaic. Each channel is adjusted to its a-priori average value. Therefore mosaic-wide average relationships between image channel values are maintained. Summary: The following describes the behavior of radiometric adjustments when operating on multi-channel imagery. HotSpot: Each channel is adjusted independently based on a mathematical model of the 'hot-spot' effect. Each channel's intensity value is adjusted to match that channel's own target value. ImageDodging: Each channel is adjusted independently based on an abstract mathematical model designed to produce uniform intensity. Each channel's intensity value is adjusted to match that channel's own target value. GlobalTilting: Each channel is adjusted independently for each image to best match the corresponding channel of the neighboring images. Mosaicking: During all mosaicking operations (PlainMosaicing, SheetCentered, AdaptiveFeathering) all image channels are blended by the same amounts at the same locations. Therefore, the relative band-to-band intensity values are preserved across transitions from one component image to another. File format support: Currently supported multi-channel image formats include the TIFF variants and also BIP/BIL/BSQ formats
6.1 Multi-Channel Image Support
6.2 Batch Mode Processing Capabilities

The following options control startup behavior and are useful for batch processing. -project=projectFileName -config=configFileName -batch
Page 69
The project file is the same as the project file saved by OrthoVista. The config file is the same format as the OrthoVista "orthovista.cfg" configuration file. The precedence of config file use is: No arguments: Use default file (orthovista.cfg) on the directory C:\Documents and settings\All Users\Application Data\Inpho\OrthoVista4\Settings (on Windows 2000 and Windows XP) respectively C:\ProgramData\Inpho\OrthoVista4\Settings (on Windows Vista). The -config argument: Use specified file in place of default The project and config files can be created (in advance of processing) by interactively defining an OrthoVista project then saving the results into an OrthoVista project and config file format. The config file of version 4.5 upward defines non-project related parameters like the cache size settings, parallel processing on/off and number of parallel jobs and the log file verbose. All project related settings are stored in the new project file format ipd. OrthoVista version 4.5 still supports project files and configuration files of older versions (ovd, cfg). If these files are given, OrthoVista 4.5 treats them as they were treated with former versions.
6.2.1 Examples
Preload Specific Project File The following will start OrthoVista and load input data from the file 'myproject.ipd' OrthoVista -project=myproject.ipd Batch Mode Processing The following will run OrthoVista in batch mode. The input data (e.g. source images, and area to be processed) will be read from the project file 'myarea.ipd'. The global settings like the Cache Size will be read from the configuration file 'myoptions.cfg'. If the cfg file is not defined the default orthovista.cfg file is used to define necessary global settings. OrthoVista -batch -project=myprj.ipd OrthoVista -batch -project=myprj.ipd -config=myoptions.cfg
6.3 Using user-defined vectors

Exclusion areas can be used to avoid using data from anomalous areas during radiometric processing. For example, it may be desirable to avoid using pixels within large water bodies during processing. During radiometric adjustments, OrthoVista determines the intensity (and color) of the input images by sampling the images in a two dimensional pattern which extends to cover all images. If a large area represents a non-characteristic intensity or color (e.g., large water bodies), the samples taken from this area can
Page 70
bias the computed solution from what is desired. For example, a bay may span many images. Some images may have very little land area covering only an edge or a corner of the image. Other images may show no land at all. The exclusion area polygon file has to be selected in the processing options dialog for Hot Spot Removal and is also only applied with Hot Spot Removal. OrthoVista allows display of the exclusion polygons as described below: 1. Select Load Region Definition from the User Data menu. OrthoVista displays the Open dialog. 2. Locate and select your vector data file and click OK. OrthoVista displays the vectors in magenta. If you dont want to display the excluded area, then clear the vectors. 1. Select Clear User Data Layers from the User Data menu. The user vectors will disappear from the screen.
6.4 Radiometric Models

Under single image balancing options in both Hot Spot Removal and Intensity Dodging options (see the following two sections respectively), there are two methods for balancing image intensity. OrthoVista supports either additive or multiplicative methods as options for intensity balancing. These are two models by which the scene intensity (signal) and unwanted intensity (noise) may be combined to produce the observed image (measurement). Conceptually, the balancing model can be thought of as mathematical function that represents the image intensity as a function of position within the frame (e.g., I(x,y)). The intensity observed at the image is a combination of two functions: the desired scene intensity (e.g., S(x,y)) and the undesirable intensity variation (e.g., N(x,y)). The job of intensity correction is to remove the N(x,y) component from I(x,y) so that only S(x,y) remains. The first is the Additive model, viz.: I(x,y) = S(x,y) + N(x,y) The second is the Multiplicative model, viz.: I(x,y) = S(x,y)*N(x,y) Note that use of a logarithmically sensitive detector (e.g., film density values) effectively transforms a multiplicative model into an additive one. Application of the additive model produces results in which intensity values are shifted or offset from those of the input imagery. Therefore, the additive model affects primarily image intensity values. For example, consider only a small image area (so that the (x,y) dependency can be neglected). The result of using the additive correction would then tend to shift the intensity histogram for this area to the left or the right. It would tend to leave the shape unchanged (other than end clipping effects). Note however, that the histogram shape for the entire image (not shown) is affected because of the spatial dependency over large image areas.
Page 71
Additive Balancing: Affects primarily image brightness
Input Histogram
Additive Offset
Figure 54: Additive Correction Method Application of the multiplicative model modifies input intensity values with both a gain and an offset. Therefore, the multiplicative model affects both image intensity and image contrast. For example, considering the same small image area as in the above example, application of the multiplicative method would tend to stretch or compress the histogram in addition to shifting it to the left or the right.
Multiplicative Balancing: Affects both brightness and contrast
Input Histogram
Multiplicative Gain
Figure 55: Multiplicative Correction Method Although it is possible to select a preferred model based on analysis of data lineage and theoretical considerations, it is generally more pragmatic to try each method on data sets representative of an application. Then, use the model that works best. In addition to selecting the method that works best for you, another practical consideration is the amount of processing time required for each method. Since the Multiplicative method is an iterative procedure, it can require considerably more processing time than the Additive method, typically increasing processing time by a factor of 3-to-5. A maximum value can be set on the number of iterations for the multiplicative method.
Page 72
6.5 Hot spot removal

The hot spot removal method for single image processing consists of a mathematical model for handling image "hot spots" common to aerial photography. This model is based on physical laws representing scene illumination and light propagation. To provide maximum utility under a variety of actual conditions and to mitigate sensitivity to special cases, the physical model is simplified to its fundamental components. The locations at which the image intensity is desired to be a specific value are distributed across the image in a grid pattern. The number of grid elements is controlled by the Grid Sampling Size parameter. The value to which the image intensity is transformed is either the explicitly provided value or the computed image average value. The choice depends on the setting of the Maintain Average Input Intensity option. When selected as the balancing method, hot spot removal is actually implemented in two components. The first is the intensity model as just described. The second component is a color model. The color model is decoupled from the intensity model and can be independently enabled or disabled via the Apply Color Correction checkbox. When Apply Color Correction is enabled, the image color is adjusted with a loworder detrending algorithm. This is useful to compensate for strange color variation within an image (e.g., compensate for color change during the course film scanning as the scanner warms, etc.). Considerations: Pros: Based on physical lighting model Allows color trend removal Cons: Can be sensitive to high-order departures from simple model (e.g., water glint, changing soil types, etc.) Can overreact near image corners and edges The Excluded Area section allows you to use ArcView shape files or DXF files to exclude an area from balancing operations. Either type in the path and filename of the shapefile, or use the Browse button to select the desired file and exclude the area inside the selected vectors from image balancing operations.
6.6 Intensity Dodging

Intensity Dodging is one of the single image or balancing adjustments. As the name implies, the intensity dodging method attempts to flatten the intensity of an image by suppressing large-scale variations. This balancing method does not affect image color; only the intensity is adjusted. Intensity dodging is conceptually similar to the dodging technique applied in photographic darkrooms. In the photo lab case, a technician modifies the image by lightening dark areas and darkening light ones. This is essentially how the intensity-dodging algorithm works, at a level of detail controlled by the Grid Size parameter. This method is an abstract algorithm and does not have a physical interpretation. The dodging method utilizes two-dimensional mathematical equations. These equations modify the image intensity so that it takes on a specified value at
Page 73
locations arranged on a regular grid. The intensity dodging method can be applied in one of two modes, additive or multiplicative. The locations at which the intensity is forced to a specific value are distributed across the image in a grid pattern. The number of grid elements is controlled by the Grid Size setting in the Dodging options dialog. The larger the number of grid elements, the more uniform the intensity of the resulting image. The value to which the image intensity is transformed is either the explicitly provided value or the computed image average value. The choice depends on the setting of the Maintain Average Input Intensity option. Intensity values used in the adjustment are computed using pixel neighborhood average values computed at each of the evaluation grid positions. The size of the pixel neighborhood is determined by the Sample Size option. With a very large number of grid values, the resulting image will have a fairly flat or uniform intensity. For example, a grid size of 2 implies one grid location at each corner of the image. Therefore, after adjustment (using grid size of 2) each of the corners should exhibit nearly the same average intensity. Note that this method is most appropriate if the image scene content is relatively homogeneous across the images. However, if image scene content changes dramatically, this method may overcorrect image intensity in some areas. Considerations: Pros: Is conceptually similar to dodging performed in a photographic darkroom Capable of modeling complex intensity variations Very stable computation across entire image Cons: No physical basis for this model Does not provide color trend removal
6.7 Coordinate reference

The reference of the coordinates in the georeference files concerning the pixel position depends on the georeference definition. In all georeference files, but TiffWorld and GeoTIFF, the reference is the outer border of the pixel (lower left corner, upper right corner ...). In TiffWorld the reference is the center of the upper left pixel in the upper left corner. GeoTIFF allows different definitions. This definition is stored with the coordinates in the GeoTIFF header.
6.8 Non constant pixel size, odd offset of orthophotos and tiles
OrthoVista can handle only Orthophotos with quadratic pixels having the same size in x and y. If Orthophotos have non-quadratic pixels then OrthoVista will compute a mean value for the pixel size and treats the orthophoto like the pixel size would have the mean value.
Page 74
Orthophotos which do not have an odd offset If Orthophotos do have an odd offset then OrthoVista shifts the orthophoto in the ground space which leads then to a scaling problem Tiles which do not have an odd offset against the orthophotos If the mosaic has a odd offset then OrthoVista shifts the orthophotos so that the tile pixels and orthophoto pixels fit with each other
If OrthoVista would fully accept the above mentioned cases then OrthoVista would have to resample the orthophotos. A resampling of orthophotos would have the negative effect that the quality of the orthophotos would degrade and the processing time would increase. You will never get problems with odd offsets if you make sure with your orthophoto generation that the coordinates of the upper left corners of the upper left pixel are dividable with whole-numbers by the pixel size. Also the coordinates of the upper left corner of your tiles should be dividable with whole-numbers by the pixel size. Note: The tfw files define the pixel center of the upper left corner. Therefore if you check the coordinates of the orthophotos then subtract pixel size in x and add pixel size in y to get the upper left corner of the upper left pixel. Therefore, we very much recommend that when you calculate the orthophotos, that you define corner coordinates that are dividable with whole-numbers by the pixel size. In addition to this the mosaic corner coordinates should also be dividable with whole-numbers by the pixel size. Example of non supported situations:
Figure 56: Orthophotos having an odd offset
Figure 57: Tiles with odd offset against. Orthophotos with no odd offset.
Page 75
6.9 Processing large blocks with 32bit version of OrthoVista

With the introduction of OrthoVista V4.0.0, the memory management of OrthoVista was changed completely. From this version on, the meta data is no more kept completely in memory but stored on the disk if not currently needed, and if the given Cache Size overruns. With this technique OrthoVista should be now able to handle any size of a block and any number of images. But still with this version, it might be necessary to subdivide large blocks in sub-blocks. Not enough disk space available A block part has to be delivered before another one Processing time for the whole block lasts too long Especially with the Windows 2000 and XP 32bit Vista 32bit operating systems we detected another technical problem, called Memory Fragmentation. Like disk space the memory can become fragmented during the processing. This then leads to a problem that even with the new memory technique, not enough RAM is available to process large blocks. Can even happen with 2GB of RAM, and requires the block has to be subdivided to process properly. If you have to run large blocks we very much suggest installing as much RAM as possible in the computer (2GB on a standard Windows operating system. 3 GB and more if you run the Windows 2000 Advanced Server Edition or a 64bit operating system). If you have problems running a large block, we suggest two methods to successfully complete the process. For large blocks you should not enable the option Save Vector Seams.
6.9.1 Two step processing

This method can be used if the RAM overruns. The Radiometric processing (apply Radiometrix changes, apply single image adjustments, apply global images adjustments) can be done in a first step and the Adaptive Feathering can be done in a second step. To do this, load all images, select all images, set your processing options, select Save Adjusted Images and switch off Save Mosaic output and process the images. Load then in a second step all radiometrically adjusted images set Single and Group Image Adjustment to None, switch off Save Adjusted Images and switch on Save Mosaic output and process the mosaics. Advantage of this method: Not so much memory is needed to process a block Steps 1 and 2 are leading faster to a result. Disadvantage of this method: More disk space to save the adjusted images needed Process has to be started a second time interactively. Whereas this disadvantage can be avoided by using the batch processing functionality There is no guaranty that employing this method will not lead to a memory overrun because of fragmentation problems
Page 76
6.9.2 Subdividing a block in sub-blocks

If you have to subdivide a block in sub-blocks, you will face two problems. The first problem is that the radiometry of the single sub-blocks might be different resulting in final mosaics that do not fit together as well. Overlapping the sub-blocks by one image and using the meta data files can reduce this problem. By doing this sub-blocks are bridged via the meta data, and the mosaics along the sub-blocks will radiometrically fit. The second problem is that the seam lines of the Adaptive Feathering function will not fit together, and you could have geometric problems along the sub-blocks like shifted buildings. Defining the sub-blocks so that they do not intersect developed areas can reduce this problem. If this is not possible, you will have to use the Seam Editor to correct seams in such situations.
6.10 Processing large blocks with 64bit version of OrthoVista

With the introduction of 64bit systems, OrthoVista is now able to handle any size of a block and any number of images. For large projects, the workstation should be equipped with at least 4GB RAM.
6.11 Processing speed

The processing speed depends very much on following facts: Size of the block Available memory of the computer and computer speed Orthophoto overlap: Should not be too large. 10 - 20% is optimal. Orthophotos are rotated against each other: OrthoVista must resample the images several times, which is very time consuming. Please rotate such images once beforehand. (Select "Rotate adjusted images before saving") Orthophotos are rotated against the mosaic area OrthoVista must resample the images several times, which is very time consuming. Please rotate such images once beforehand. (Select "Rotate adjusted images before saving") Multiplicative - additive method The multiplicative method works iterative and therefore needs more time than the additive method. Hot Spot Removal Sampling Grid Size Adaptive Feathering Alpha Grid Size Scan resolution and color images Larger images have more pixels requiring more processing time. In addition color images have to be transformed several times into different color spaces requiring more processing time than B&W images. Overviews. Each image should have at least one overview. The overview speeds up the image display very much but also the processing of data. A
Page 77
full set of overviews is advisable in case you work interactively with the images in OrthoVista but especially in Seam Editor. Number of iterations set for Global Tilting Adjustment Contrast adjustment on/off The color adjustment should be always enabled. But the contrast adjustment should be only done if necessary. RAM. Your computer should be equipped with at least 1 GB of RAM. If your computer does not have 2 GB of RAM, it is possible to have physical RAM overruns and Virtual Memory has to be used. In this case, memory will be swapped to the disk by the operating system. This leads to a slower processing time. Cache Size For 32bit operating systems, we recommend setting the cache size to 128 MB. According to our testing a cache size of 256 MB does not speed up the processing of normal images but you run into the risk that the physical memory overruns or in case the memory gets too fragmented the memory at all overruns and OrthoVista cant finish the job. Whereas in case of very large images e.g. ADS40 images we very much suggest setting the cache size to 256 MB. For 64bit operating system a cache size of 512 or 1024 can be used if the workstation is equipped with at least 4GB RAM. Computer type Vista 64bit or Server 2008 is in general about 20% faster in file IO than the XP operating systems. Enabling Sub-Processing When sub-processed are used for writing images then writing of the images is about 2 times or even more faster. Writing images might need a big portion of the overall OrthoVista processing time.
Page 78
End User License Agreement

This INPHO End User License Agreement, hereinafter referred to as EULA, is a legal Agreement between you, the end user, hereinafter referred to as "END USER" and INPHO GmbH, Stuttgart, Germany, hereinafter referred to as "INPHO" for the software products of INPHO, hereinafter referred to as "SOFTWARE PRODUCT(S)". By installing, copying, or otherwise using SOFTWARE PRODUCT(S), END USER agrees to be bound by the terms of this EULA. INPHO agrees to grant hereby, and END USER agrees to accept a nontransferable, and non-exclusive licence to use SOFTWARE PRODUCT(S) of INPHO under the following terms and conditions. 1. END USER must have acquired from INPHO or from an authorized distribution partner of INPHO, a license for the use of SOFTWARE PRODUCT(S). 2. SOFTWARE PRODUCT(S) shall be licensed under this EULA effective from the date of installation in decoded form. This Agreement shall remain in force until END USER discontinues the use of SOFTWARE PRODUCT(S) and returns SOFTWARE PRODUCT(S) to INPHO, or this EULA is otherwise terminated as provided herein. 3. END USER shall have the right to use SOFTWARE PRODUCT(S) or any portion thereof at any given time on one (1) single computer (CPU) only. In case of transfer of SOFTWARE PRODUCT(S) to another computer, END USER guarantees and ensures to make use of SOFTWARE PRODUCT(S) on one (1) single computer (CPU) only at any given time. 4. END USER is entitled to install SOFTWARE PRODUCT(S) on as many computers as he wishes if and only if END USER guarantees and ensures to make use of SOFTWARE PRODUCT(S) on one (1) single computer (CPU) only at any given time. END USER may make backup copies of SOFTWARE PRODUCT(S). 5. END USER agrees that he and his employees will not make available SOFTWARE PRODUCT(S), or portions thereof, to any persons other than END USER or his employees. END USER is liable that the SOFTWARE PRODUCT(S) or any portion thereof, will not be copied except as necessary for use under this EULA. 6. In case of educational licenses of SOFTWARE PRODUCT(S), END USER agrees that he or his employees or his students will use SOFTWARE PRODUCT(S), or portions thereof, for educational and research purposes only and will not use SOFTWARE PRODUCT(S), or portions thereof, for any commercial purposes. END USER is liable that educational licenses of SOFTWARE PRODUCT(S) or any portion thereof, are used for the agreed purposes only.
Page 79
7. INPHO warrants that it is the owner of SOFTWARE PRODUCT(S) or that it has the right to grant a sub-licence to use SOFTWARE PRODUCT(S). Title and ownership of SOFTWARE PRODUCT(S) shall at all times remain with INPHO or relevant third parties. 8. INPHO warrants for a period of 12 months from the date of delivery that, under normal use, SOFTWARE PRODUCT(S) will perform substantially in the manner described in the documentation. The responsibility of INPHO for quality and performance of SOFTWARE PRODUCT(S) ceases with the end of the warranty period. 9. INPHO does not warrant that SOFTWARE PRODUCT(S) will meet END USER's requirements and that SOFTWARE PRODUCT(S) will operate error-free. 10. INPHO's entire liability shall be either the repair or the replacement of any SOFTWARE PRODUCT(S), which does not meet the respective warranties given above, or a refund of the purchase price of the defective software. 11. In no event will INPHO be liable for any damages arising out of or in connection with the use of SOFTWARE PRODUCT(S) or the inability to use SOFTWARE PRODUCT(S). 12. In the event END USER neglects or fails to pay the appropriate purchase price or licence fee(s), or to adhere to any of its obligations hereunder, this EULA shall immediately terminate, and SOFTWARE PRODUCT(S) will have to be returned to INPHO immediately. 13. All additional arrangements or alterations of this Agreement require the written permission of both parties. The laws of the Federal Republic of Germany shall govern this EULA. The place of jurisdiction is Stuttgart, Germany.
Page 80

ReferenceManual OrthoVista (English) 45

Uploaded by

Copyright:

Available Formats

ReferenceManual OrthoVista (English) 45

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

ReferenceManual OrthoVista (English) 45

Uploaded by

Copyright:

Available Formats

OrthoVista 4.

Installing OrthoVista .................................................................. 5

Getting Started with OrthoVista................................................ 7

Shortcuts................................................................................................... 12 Plugins ...................................................................................................... 12

4.3.1 4.3.2 4.3.3

4.4 4.5 4.6 4.7

Loading input image data........................................................................ 20 General Preferences ................................................................................ 22

4.10 4.11 4.12 4.13 4.14 4.15 4.16 4.17 4.18

Radiometrix Editor ................................................................................... 31

Color Picker .............................................................................................. 40 Status Dialog ............................................................................................ 40

6.3 6.4 6.5 6.6 6.7 6.8 6.9

Processing large blocks with 64bit version of OrthoVista................... 77 Processing speed..................................................................................... 77

End User License Agreement ................................................. 79

Reference Manual OrthoVista

1.1 Quick start

1.2 What is OrthoVista?

1.3 What can OrthoVista do for you?

Reference Manual OrthoVista

Sample data set before OrthoVista processing.

1.3.1 Manual or semi-automatic adjustments

1.3.2 Balancing and tilting

Reference Manual OrthoVista

Sample data set after OrthoVista processing

1.4 How does OrthoVista fit into your workflow?

1.5 Will OrthoVista solve all your imagery problems?

Reference Manual OrthoVista

Reference Manual OrthoVista

2.1 What are the system requirements for running OrthoVista?

2.2 Linux installation

2.3 Windows installation

2.3.1 Windows registration

2.3.2 Linux registration

Reference Manual OrthoVista

Reference Manual OrthoVista

Getting Started with OrthoVista

3.1 Invoking OrthoVista

The OrthoVista Main Window

Reference Manual OrthoVista

3.2 Exiting OrthoVista

Reference Manual OrthoVista

4.1 Main window

Display Layer Popup Menu

Reference Manual OrthoVista

4.1.1 Show Histogram

Selection of an image to be on top

Reference Manual OrthoVista

4.1.3 Display Layers

4.1.4 Viewing Control

Reference Manual OrthoVista

4.3.1 Image support plugins (image formats)

4.3.2 Geo-reference data support plugins (report formats)

Reference Manual OrthoVista

4.3.3 Image adjustment plugins (image processing)

4.4 Rotated images

Reference Manual OrthoVista

4.5 File management

Reference Manual OrthoVista

4.6 Project Dialog

Reference Manual OrthoVista

4.7 Defining the processing area