Image Processing

Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 8

Image Processing

Divya M.R
IT Branch

Abstract stochastic models. It is represented by AR model.


In this paper, the basics of capturing an image, Degradation is represented by MA model. Other
image processing to modify and enhance the form is orthogonal series expansion. Image
images are discussed. There are many processing system is typically non-casual system.
applications for Image Processing like Image processing is two dimensional signal
surveillance, navigation, and robotics. Robotics processing. Due to linearity Property, we can
is a very interesting field and promises future operate on rows and columns separately. Image
development so it is chosen as an example to processing is vastly being implemented by “Vision
explain the various aspects involved in Image Systems” in robotics. Robots are designed, and
Processing . The various techniques of Image meant to be controlled by a computer or similar
Processing are explained briefly and the devices. While “Vision Systems” are most
advantages and disadvantages are listed. There sophisticated sensors used in Robotics. They relate
are countless different routines that can be the function of a robot to its environment as all
used for variety of purposes. Most of these other sensors do. “Vision Systems” may be used
routines are created for specific operations and for a variety of applications, including
applications. However, certain fundamental manufacturing, navigation and surveillance.
techniques such as convolution masks can be Brief History
applied to many classes of routines. We have One of the first applications of image processing
concentrated on these techniques, which enable techniques in the first category was in improving
us to adapt, develop, and use other routines and digitized newspaper pictures sent by submarine
techniques for other applications. The advances cable between London and Newyork . Introduction
in technology have created tremendous of the Bartlane cable picture transmission system in
opportunities for visual system and image the early 1920’s reduced the time required to
processing. There is no doubt that the trend will transport a picture across the Atlantic from more
continue into the future. than a week. Some of the initial problems in
improving the visual quality of these early digital
Keywords pictures were related to the selection of printing
Image, ImageProcessing , Surveillance, procedures and the distribution of brightness levels.
Robotics, Navigation, Convolution. The printing method used to send the picture by
specialized printing equipment coded pictures
Introduction was abandoned towards the end of 1921 in favor of
a technique based on photographic reproduction
Digital Image Processing is concerned with
made from tapes perforated at the telegraph
acquiring and processing of an image. In simple
receiving terminal whose improvements are
words an image is a representation of a real scene,
evident both in tonal quality and resolution. During
either in black and white or in color, and either in
this period introduction of a system for developing
print form or in a digital form i.e., technically a
a film plate via light beams that were modulated by
image is a two-dimensional light intensity function.
the coded picture tape improved the reproduction
In other words it is a data intensity values arranged
process considerably. Improvements on processing
in a two-dimensional form like an array, the
methods for transmitted digital pictures continued
required property of an image can be extracted
to be made during the next 35 years. However, it
from processing an Image .Image is typically by
took the combined advents of large scale digital
computers and space program to bring into focus image quality for human perception and / or
the potential of image processing concepts. Work computer interpretation. Image Enhancement. To
on using computer techniques for improving bring out detail is obscured, or simply to highlight
images from a space probe began at the certain features of interest in an image.
Jetpropulsion Laboratory at Pasadena, California in Example:
1964 when pictures of the moon were transmitted Image
which were processed by computer to correct
various types of image distortion inherent in the
on-board television camera. From 1964 until the
present, the field of image processing has grown
vigorously. The second major area of application of
digital image processing techniques is in solving
problems dealing with machine perception. In this
case, interest focuses on procedures for extracting
from an image information in a form suitable for Image Restoration:
computer processing. Often, this information bears Improving the appearance of an image tend to be
little resemblance to visual features that human based on mathematical or probabilistic models of
beings use in interpreting the contents of an image. image degradation.
Examples of the type of information used in Example:
machine perception are Statistical moments,
Fourier transform coefficients, and multi
dimensional distance measures.
Index Terms
Image:
An image may be defined as a two-dimensional
function, f(x, y), where x and y are spatial (plane) Distorted Image Restorted
Image
coordinates, and the amplitude of ‘f’ at any pair of
coordinates (x, y) is called the intensity or gray
Color Image Processing:
level of the image at that point. When x, y and the
Gaining its importance because of the significant
amplitude values of f are all finite, discrete
increase in the use of digital images over the
quantities, we call the image a digital image. The
Internet.
field of digital image processing refers to
processing digital images by means of a digital
Wavelets:
computer. A digital image is composed of a finite
Foundation for representing images in various
number of elements, each of which has a particular
degrees of resolution. Used in image data
location and value. These elements are referred to
compression and pyramidal representation (images
as picture elements, image elements, pels, and
are subdivided successively into smaller regions)
pixels. Pixel is the term most widely used to denote
the elements of a digital image Compression:
Reducing the storage requires to save an image or
the bandwidth required to transmit it. Ex. JPEG
(Joint Photographic Experts Group) image
compression standard. Other areas of image
processing technology include:
o 3-D Imaging
o Multimedia
o In Camera Image Processing

Image Processing Inside the Digital


Camera:
Image Processing: "You press the button. We do the rest" takes on
Image processing is a subclass of signal processing new meaning when you consider that the moment
concerned specifically with pictures. Improve you press the button of a digital camera, several
things happen almost instantaneously to produce
bright colorful images. When you snap a picture
with a digital camera, sophisticated digital image
processing takes place inside the camera.
1. Light is captured on an image sensor with pixels
coated in red, green and blue patterns. The sensor
captures the light and interpolates the scene into
full color.
2. Smart algorithms in the camera "firmware"
respond to the scene data and sharpen the image.
3. The system automatically corrects color, adjusts Image Segmentation:
for exposure and implements automatic noise Computer tries to separate objects separate objects
reduction. from the image background from the image
4. Finally, the camera compresses the image into background. It is one of the most difficult tasks in
the JPEG format. DIP. A rugged segmentation procedure brings the
process a long way toward successful solution of
3D Imaging
an image problem Output of the segmentation stage
The world about us is three-dimensional, yet
is raw pixel data, constituting either the boundary
photography shows us the world in two
of a region or all the points in the region itself.
dimensions.Three-dimensional imaging attempts to
retain the three-dimensional nature of the world.
Scientists are working on cameras and applications
that collect and apply the full geometry of world.
They blend computer graphics and photography so
objects can be viewed from any vantage and still
have the detail and richness of photographs. While
our methods initially are intended for e-commerce,
at some point, the technology will be available to
consumers.

Analysis:
The following is the overall view and analysis of
Image Processing.
Image Processing Techniques:
Image Processing techniques are used to enhance,
improve, or otherwise alter an image and to prepare
it for image analysis. Usually, during image
Key Research Areas: processing information is not extracted from the
o Photography image. The intention is to remove faults, trivial
o Software information, or information that may be important,
o Computer Graphics but not useful, and to improve the image. Image
o Displays processing is divided into many sub processes,
Recent Achievements: including histogram Analysis, Thresholding,
Developed with a system that collects images of an Masking, Edge Detection, Segmentation, and
object and builds a computer graphic model that others.
can be viewed from any point. The system
eliminates much of the time and expense in
constructing 3D objects of high quality. Objects
now are photo-realistic since the data comes
directly from photograph.
Morphological processing:
Tools for extracting image components that are Stages In Image Processing:
useful in the representation and description of
shape. A problem
A solution level as a threshold, comparing each pixel value
Image with the threshold, and then assigning the pixel to
Acquisition Recognition the different portions, depending on whether the
and pixel’s grayness level is below the threshold or
Interpretation above the threshold value. Thresholding can be
Knowledge nn performed either at a single level or at multiple
Base levels, in which the image is processed by dividing
it into ”layers”, each with a selected threshold.
Various techniques are available to choose an
appropriate threshold ranging from simple routines
Representatio for binary images to sophisticated techniques for
Preprocessin
Segmentatio n and complicated images.
g
n Description Connectivity:
Sometimes we need to decide whether neighboring
pixels are somehow “connected” or related to each
other. Connectivity establishes whether they have
the same property, such as being of the same
region, coming from the same object, having a
similar texture, etc. To establish the connectivity of
1. Image Acquisition: neighboring pixels, we first have to decide upon a
An image is captured by a sensor (such as a connectivity path.
monochrome or color TV camera) and digitized. If Noise Reduction:
the output of the camera or sensor is not already in Like other signal processing mediums, Vision
digital form, an analog-to digital converter digitizes Systems contains noises. Some noises are
it. systematic and come from dirty lenses, faulty
2. Recognition And Interpretation: electronic components, bad memory chips and low
Recognition is the process that assigns a label to an resolution. Others are random and are caused by
object based on the information provided by its environmental effects or bad lighting. The net
descriptors. effect is a corrupted image that needs to be
Interpretation is assigning meaning to an ensemble preprocessed to reduce or eliminate the noise. In
of recognized objects. addition, sometimes images are not of good
3. Segmentation: quality, due to both hardware and software
inadequacies; thus, they have to be enhanced and
Segmentation is the generic name for a number of
improved before other analysis can be performed
different techniques that divide the image into
on them.
segments of its constituents. The purpose of
segmentation is to separate the information Convolution Masks:
contained in the image into smaller entities that can A mask may be used for many different purposes,
be used for other purposes including filtering operations and noise reduction.
4. Representation And Description: Noise and Edges produces higher frequencies in
the spectrum of a signal. It is possible to create
Representation and Description transforms raw
masks that behave like a low pass filter, such that
data into a form suitable for the Recognition
higher frequencies of an image are attenuated while
processing.
the lower frequencies are not changed very much.
There by the noise is reduced.

5. Knowledge Base:
Edge Detection:
A problem domain detailing the regions of an
image where the information of interest is Edge Detection is a general name for a class of
routines and techniques that operate on an image
known to be located is known as knowledge base.
and results in a line drawing of the image. The
It helps to limit the search.
lines represented changes in values such as cross
Thresholding:
sections of planes, intersections of planes, textures,
Thresholding is the process of dividing an image
lines, and colors, as well as differences in shading
into different portions by picking a certain grayness
and textures. Some techniques are mathematically
oriented, some are heuristic, and some are of system performance models. For hybrid and
descriptive. All generally operate on the digital imaging products, modeling is used to
differences between the gray levels of pixels or predict the performance the entire system,
groups of pixels through masks or thresholds. The including elements such as optics, image sensors,
final result is a line drawing or similar scanners, image processing operations, printers,
representation that requires much less memory to emissive displays, capture and display media and
be stored, is much simpler to be processed, and human visual responses. In a digital camera, for
saves in computation and storage costs. Edge example, modification of only one component --
detection is also necessary in subsequent process, such as the lens, sensor, or color filter array -- can
such as segmentation and object recognition. significantly impact image quality.
Image Data Compression: Image Simulation:
Electronic images contain large amounts of Image simulation is used to generate actual images
information and thus require data transmission based on system models. Scientists use advanced
lines with large bandwidth capacity. The imaging software to emulate the characteristics of
requirements for the temporal and spatial resolution film or digital systems. They manipulate tone scale,
of an image, the number of images per second, and color, sharpness and grain to generate the kind of
the number of gray levels are determined by the picture quality that a proposed system would
required quality of the image. Recent data deliver. The images generated are then used in
transmission and storage techniques have studies in which people choose which pictures they
significantly improved image transmission find most pleasing. Scientists work on simulating
capabilities, including transmission over the traditional and digital imaging systems, as well as
Internet. hybrid systems that combine both technologies.
Real-Time Image Processing: Drawing on an extensive library of fully
In many of the techniques considered so far, the characterized images, these scientists are able to
image is digitized and stored before processing. In save considerable time and effort in the process.
other situations, although the image is not stored, Simulating new color films and paper emulsions
the processing routines require long computational has replaced time-consuming, expensive physical
times before they are finished. This means that, in experiments in chemical laboratories. In digital
general, there is a long lapse between the time and imaging, simulations can eliminate the need for
image is taken and the time a result obtained. This elaborate "breadboard" hardware needed to
may be acceptable in situations in which the produce experimental images.
decisions do not affect the process. However, in Image communications:
other situations, there is a need for real-time Image communications is what makes digital
processing such that the results are available in real imaging much more than a mere replacement for
time or in a short enough time to be considered real film-based technology. Advances in information
time. Two different approaches are considered for networking allow applications to seamlessly
real time processing. One is to design dedicated integrate multimedia into every form of
hardware such that the processing is fast communication.
enough to occur in real time. The other is to try to  How to Manage and Use All That
increase the efficiency of both the software and the Data
hardware and thereby reduce processing and The digitization of the imaging industry brings new
computational requirements. technical challenges and opportunities.The vision
Modeling: of helping people and businesses tell their stories
A key goal for the system designers was for overall more effectively through imaging, once viewed as
image quality to closely match that of current revolutionary in the imaging industry, is now taken
35mm systems. Developers used predictive system for granted.
modeling to meet image quality targets for the  Making Digital Imaging Easy Is Not
system. Image quality modeling was key in
So Easy
selecting the new film and print format sizes and in
As camcorders, digital still cameras, and PDAs
evaluating proposed features, such a pseudo zoom.
converge — and as more powerful editing and
Ultimately, specifications for such elements as film
organization tools are developed — we will enter a
design, camera exposure control and component
new age of personal imaging in which anyone
positioning tolerances were developed with the aid
could produce multimedia content that rivals the usage of sensors/cameras and Edge Detection
quality of today's movies and television programs. technique are related to Image Processing and
Achieving this grand vision will be neither quick Vision Systems. By this technique the complexity
nor easy. Today, image communication specialists of using manual sensors is minimized to a great
are creating the infrastructure that could one-day extent and thereby sophistication is increased.
support this vision, as well as the technologies that Hence image processing is used here in the study
will make that possible. These efforts should give a of robotics.
sustainable, competitive edge in the new age of Medical imaging
imaging. Increasingly we are discovering that Medical imaging refers to the techniques and
XML offers a common framework that can solve processes used to create images of the human body
many of the problems that arise from e-commerce, (or parts thereof) for clinical purposes (medical
privacy of records, and multiple data formats. Here procedures seeking to reveal, diagnose or examine
we want to present some of the applications of disease) or medical science (including the study of
Image Processing in some fields where it is applied normal anatomy and physiology). As a discipline
like Robotics, Medical field and common uses… and in its widest sense, it is part of biological
Applications: imaging and incorporates radiology (in the wider
There is a wide range of applications of the digital sense), adiological sciences, endoscopy,(medical)
image processing, few f them are thermography, medical photography and
1. Computer vision microscopy (e.g. for human pathological
2. Face detection investigations).Measurement and recording
3. Feature detection techniques which are not primarily designed to
4. Lane departure warning system produce images, such as
5. Non-photorealistic rendering electroencephalography(EEG) and
6. Medical image processing magnetoencephalography (MEG) and others, but
7. Microscope image processing which produce data susceptible to be represented as
8. Morphological image processing maps (i.e. containing positional information), can
Robotics: be seen as forms of medical imaging.
Image Processing is vastly being implemented in Electron microscopy
Vision Systems in Robotics. Robots capture the The electron microscope is a microscope that can
real time images using cameras and process them magnify very small details with high resolving
to fulfill the desired action. A simple application in power due to the use of electrons as the source of
robotics using Vision Systems is a robot hand-eye illumination, magnifying at levels up to 2,000,000
coordination system. Consider that the robot’s task times. Electron microscopy is employed in
is to move an object from one point to another anatomic pathology to identify organelles within
point. the cells. Its usefulness has been greatly reduced by
Here the robots are fixed with cameras to view the immune histochemistry but it is still irreplaceable
object, which is to be moved. The hand of the robot for the diagnosis of kidney disease, identification
and the object that is to be captured are observed of immotile cilia syndrome and many other tasks
by the cameras, which are fixed to the robot in Fluoroscopy
position, this real time image is processed by the
Fluoroscopy produces real-time images of internal
image processing techniques to get the actual
structures of the body in a similar fashion to
distance between the hand and the object. Here the
radiography, but employs a constant input of x
base wheel of the robot’s hand is rotated through
rays, at a lower dose rate. Contrast media, such as
an angle, which is proportional to the actual
barium, iodine, and air are used to visualize
distance between hand and the object. Here a point
internal organs as they work. Fluoroscopy is also
in the target is obtained by using the Edge
used in image-guided procedures when constant
Detection Technique. The operation to be
feedback during a procedure is required. An image
performed is controlled by the micro-controller,
receptor is required to convert the radiation into an
which is connected to the ports of the fingers of the
image after it has passed through the area of
robot’s hand. Using the software programs the
interest. Early on this was a fluorescing screen,
operations to be performed are assigned keys from
which gave way to an Image Amplifier (IA) which
the keyboard. By pressing the relative key on the
was a large vacuum tube that had the receiving end
keyboard the hand moves appropriately. Here the
coated with cesium iodide, and a mirror at the
opposite end. Eventually the mirror was replaced
with a TV camera.
Nuclear medicine
Images from gamma cameras are used in nuclear
medicine to detect regions of biological activity
that are often associated with diseases. A short
lived isotope, such as 123I is administered to the
patient. These isotopes are more readily absorbed
by biologically active regions of the body, such as
tumors or fracture points in bones.

Baby scan MRI scan of knee

Projection radiography
Radiographs, more commonly known as x-rays, are
often used to determine the type and extent of a
fracture as well as for detecting pathological
changes in the lungs. With the use of radio-opaque
contrast media, such as barium, they can also be
used to visualize the structure of the stomach and Thyroid
intestines - this can help diagnose ulcers or certain Animal Tracking
types of colon cancer. In order to collect spatial information about an
Ultrasound animal's movement by means of digital image
Medical ultrasonography uses high frequency processing techniques the information has to be
collected sequentially. This can be achieved by
broadband sound waves in the megahertz range
analyzing subsequent image frames of a digitized
that are reflected by tissue to varying degrees to video. By means of extracting the X- and Y-
produce (up to 3D) images. This is often used to coordinates representing the position of a mouse
visualize the fetus in pregnant women. Other for each individual image frame, the path can be
important uses include imaging the abdominal measured. Ideally, time as an additional
organs, heart, male genitalia, and the veins of the information, should be included. If the images are
leg. While it may provide less anatomical detail captured and analyzed at a constant framerate or,
alternatively, the exact time for each coordinate
than techniques such as CT or MRI, it has several
pair is extracted simultaneously, valuable
advantages which make it ideal in numerous information like speed, stops, etc. can be calculated
situations, in particular that it studies the function from the data. Talking about timelines, it has to be
of moving structures in realtime, emits no ionizing noted that depending upon the framerate (time
radiation, and contains speckle that can be used in resolution) the total pathlength may vary to a great
elastography. It is very safe to use and does not extend.
appear to cause any adverse effects, Although
information on this is not well Documented. It is
also relatively inexpensive and quick to perform.
Ultrasound scanners can be taken to critically ill
patients in intensive care units, avoiding the danger
caused while moving the patient to the radiology
department. The real time moving image obtained
can be used to guide drainage and biopsy
procedures. Doppler capabilities on modern
scanners allow the blood flow in arteries and veins Advantages:
to be assessed.
1. In medicine by using the Image Processing
techniques the sophistication has increased. This
lead to technological advancement.
2. Vision Systems are flexible, inexpensive,
powerful tools that can be used with ease.
3. In Space Exploration the robots play vital role
which in turn use the image processing techniques.
4. Image Processing is used for Astronomical
Observations.
5.Also used in Remote Sensing, Geological
Surveys for detecting mineral resources etc.
6. Also used for character recognizing techniques,
inspection for abnormalities in industries.
Disadvantages:
1. A Person needs knowledge in many fields to
develop an application / or part of an application
using image processing.
2. Calculations and computations are difficult and
complicated so needs an expert in the field related.
Hence it’s unsuitable and unbeneficial to ordinary
programmers with mediocre knowledge.
Conclusion:
It’s a critical study, which plays a vital role in
modern world as it is involved with advanced use
of science and technology. The advances in
technology have created tremendous opportunities
for Vision System and Image Processing. There is
no doubt that the trend will continue into the future.
From the above discussion we can conclude that
this field has Relatively more advantages than
disadvantages and hence is very useful in varied
branches.

References:
1. Introduction To Robotics, Analysis, Systems,
Applications - Saeed B. Niku
2. Introduction To Digital Image Processing –
Anil K.Jain
3. Digital Mage Processing - Rafael C. Gonzalez
And Richard E. Woods, AddisonWesley 1993.
4. Digital Image Processing by Engr. Syed
Muhammad Munavvar Hussain.

You might also like