Uav Oblique Data and Laser Scanning in An Excavated Area: September 2016

Download as pdf or txt
Download as pdf or txt
You are on page 1of 5

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/308296235

UAV OBLIQUE DATA AND LASER SCANNING IN


AN EXCAVATED AREA

Conference Paper September 2016

CITATION READS

1 63

4 authors, including:

Antonia Spano' Giulia Sammartano


Politecnico di Torino Politecnico di Torino
81 PUBLICATIONS 173 CITATIONS 23 PUBLICATIONS 11 CITATIONS

SEE PROFILE SEE PROFILE

Lorenzo Teppati Los


Politecnico di Torino
15 PUBLICATIONS 5 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

ISPRS WG IV/2: Ontologies, Semantics, and Knowledge Representation for Geospatial Information
View project

All content following this page was uploaded by Giulia Sammartano on 19 September 2016.

The user has requested enhancement of the downloaded file.


th
Proceedings of the 8
International Congress
on Archaeology,
Computer Graphics,
Cultural Heritage and
Innovation
ARQUEOLGICA 2.0
in Valencia (Spain),
Sept. 5 7, 2016

UAV OBLIQUE DATA AND LASER SCANNING IN AN EXCAVATED AREA


Filiberto Chiabrando, Antonia Span, Giulia Sammartano, Lorenzo Teppati*
Dep. Architecture and Design, Politecnico di Torino, Viale Pier Andrea Mattioli, 39 - 10125 Torino, Italy. [email protected],
[email protected], [email protected], [email protected]

Abstract:
This paper discusses some enhancements concerning 3D modelling and integration of 3D data from aerial and terrestrial
sensors, developed by geomatics in the field of Cultural Heritage metric documentation. For archaeological purposes, it
is interesting to deal with the considerable advantages in term of sustainability (automated acquisition, quickness,
precision, time and cost cutting) of new multi-sensors approaches for the data acquisition and the management phases.
In particular, the UAV (Unmanned Aerial Vehicles) Photogrammetry with the joint use of nadiral and oblique cameras,
can be valuably combined with the large-scale details reach by terrestrial LiDAR in vast areas or complex objects,
especially in mostly vertical sized objects. Here it will be reported an experience of integrated 3D survey in an
archaeological context in Piedmont region (Italy), the Hercules Fountain in the gardens of the Venaria Reale. It has
witnessed several historical phases during centuries, from its construction in 16th to the disuse and dismantling in 17th,
up to the 21th century in which it was lastly brought back to light. The goal of the test is the generation of a 3D
continuous model of the site for documentation purposes, future consolidation and enhancement projects finalized to a
public promotion. To meet these aims a terrestrial laser scanning (TLS) survey has been designed combining terrestrial
and UAV photogrammetric data acquisition, to produce a high detailed 3D textured model from which infer standard 2D
drawings, digital orthoimages and further 3D releases. The entire workflow and outputs were compared together to
evaluate the effectiveness of each elaboration according to the survey goals.
Key words: 3D documentation, Building Archaeology, Lidar, UAVs Photogrammetry, Oblique Cameras, Savoy Architecture

complete model rich in metric and non-metric information


1. 3D data documentation role in together with the completeness provided aerial
archaeology: integration of ground- documentation by UAV. In fact, these models allow
based data and UAV nadiral and studying not only the geometry of the surveyed object,
oblique cameras via standars 2D and 3D representation (Boehler and
Marbs 2004), but also for material analysis, pathologies
New strategies with 3D multi-sensor metric survey prove and degradations, structural assessments, conservation
that the documentation process can reach good levels of monitoring (Patias 2013), virtual restoration and digital
sustainability, in the direction of quality of information communication e.g. Augmented\Virtual Reality, 3D GIS
and in terms of time and cost. The georeferencing of and HBIM tools to manage 3D reality-based information
spatial data is essential for interoperability during times, (Landeschi et al. 2016). Concurrently, UAV platforms are
as a continuous part of the whole process of monitoring becoming in the last years one of the most employed
and Heritage preservation envolving sector-specific system for aerial mapping and 3D modeling issues. For
experts. Furthermore, the attempts of effective archaeological sites, either multi-rotor or fixed wing
applications in archaeological contexts, must keep into UAVs, offer low-cost alternatives to the classical manned
account the importance of their spatial and temporal aerial photogrammetry, performing data acquisition with
complexity and the choice of targeted surveys and high-resolution digital camera in semi-automatic and
sensors in terms of scale detail requested and of autonomous ways. Point clouds, DSM/DTM, orthophoto,
resources involved. A good numbers of research in textured 3D models, 2D drawing data, etc. can be
Geomatics scenario towards these themes are quickly produced with controlled accuracy. Usually the
approaching to improve and adapt themselves to the acquisition for architectural and archaelogical
different sites needs (Remondino et al. 2011; Lerma et documentation are performed using the camera oriented
al. 2011; Balletti et al. 2015). In specific, for in the nadir direction, but a big attention is now focused
documentation purposes in the field of Building also on the use of oblique images (Rupnik et al. 2014;
Archaeology, the high scale dense data generally Xiao et al. 2012) and the research activities in the
reached by the combined use of TLS (active sensor) and geomatic field are even more oriented in optimization of
close range photogrammetry (passive sensor) lead to a these algorithms.

*
Corresponding Author: Lorenzo Teppati, [email protected]

This work is licensed under a Creative Commons 4.0 International License (CC BY-NC-ND 4.0)
EDITORIAL UNIVERSITAT POLITCNICA DE VALNCIA
350
Chiabrando, Span, Sammartano, Teppati, 2016.

2. The Hercules Fountain experience The LiDAR scans performed by Focus3D Cam2 by Faro
were planned to cover all the area n82 acquisitions at
The integration of 3D metric survey techniques is 360; n79 registered scans; 20*105 points/scans;
presented here in a complex test site: the Hercules Resolution MPti 1pt/6mm at 10m). The terrestrial
Fountain in Venaria Reale (Fig. 1), a UNESCO Heritage photogrammetric recording, using a Nikon D800E reflex
site, belonging to the residences of Savoy Royal House, digital camera (457images, 24mm lens, 36 Mpx CMOS
located 15 km northwest of Turin (Piedmont, Italy). sensor, image size 7360 x 4912 pixel, pixel 4.89 m)
Nowadays the ruins of the ancient masonries uncovered using an overlapping >80%, complete the terrestrial
11 years ago, are involved in a second step of documentation. The fieldwork was finally completed by
restoration project that needed a detailed and complete three photogrammetric flights on the area, based on
3D metric documentation, as base for restoration, nadiral and oblique acquisitions to produce a detailed
promotion and valorization projects by the Venaria Reale DSM of the fountain with a high-scale detail especially in
Royal Palace Management Consortium (client of the the vertical dimension. They were perfrmed using a
documentation). Hexakopter by Mikrokopter (Fig. 2) with a ground sample
distance of 0.51cm. In the employed configuration the
multi-rotor platform is equipped with a commercial off-
the-shelf (COTS) Sony Alpha 5100 digital camera (24.3
MPx CMOS sensor, 6000*4000 max resolution, pixel
3.92 m, equipped with a 20mm lens).

Figure 1: A view of the Reggia di Venaria Reale

Amedeo di Castellamonte (the royal architect of the


Savoy family) projected the Reggia di Venaria in the Figure 2: The multi-rotor system and the ground control station
second half of the XVII century, to be a hunting
residence for the Royal Family, and the Fontana Thanks to the Lidar precision in the very high scale, after
dErcole, richly decorated and covered with mosaics of the point clouds registrations, many orthoimages have
shells and stones, located in the lower gardens between been extracted (Using Pointcab software). Starting from
1669 and 1672 (Cornaglia 1994). The Hercules fountain this data, an accurate interpretation and drawing using
was subjected to a series of restorations and was finally CAD software (Figs. 3 y 4) have been performed.
dismantled in the 1751. The works for the restoration Furthermore, the terrestrial images collected on the field
started only in the 1999 and, after years of efforts, the were processed as well, by photogrammetric approach
original magnificence of Reggia di Venaria was returned based on Structure from motion (SfM) using Agisoft
to the public; today, with 573.337 visitors in the last year, PhotoScan software. The results from the
is one of the most visited historical sites in Italy. In the photogrammetric process on an area of 154sqm are:
summer of 2005, the Fontana dErcole was brought to (GSD) 1.23 mm/px. (RMS) on 14 GCPs: 14mm. Final
light during some work of excavation in the gardens, was point cloud, 100millions of points, contains around
restored and is now part of the tours of the Reggia 9608.06pt/m2. Final digital surface: 1.6 million of
(Cornaglia 1994). The structure of the masonries in the triangles. Finally, the workflow has included the
Hercules Fountain is quite complex and irregular; the processing of the aerial acquisitions: camera orientation,
Northern part of the structure, recently excavated, is images georefereincing using GCPs, point cloud
exposed at the ground level and constrained by many densification, mesh calculation and texturization, with the
trees, while the Southern area is still underground. software Pix4D for the generation of the final DSM and
Furthermore, the Eastern and the Western parts of the orthoimage (Fig. 3 y 5). The results from the
fountain differ in the terrain elevation and in the photogrammetric process on an area of 10.700sqm are:
upstanding structures height. Finally, all masonries (GSD) 0.51cm. (RMS) on 29 GCPs: 0.4 cm. (RMS) on
present a deep status of alteration and generally a bad 12 CPs: 0.5 cm. Final point cloud, 71 ml of points. Final
conservation, since the architectural coating is almost mesh, 10 ml of triangles.
totally lost.
In the site, an integrated multisensor 3D survey was
conducted: the UAV photogrammetric acquisitions
integrate the terrestrial one, via close-range
photogrammetry and LiDAR survey. A complete oblique
acquisition by UAV was performed with the aim to
understand the real potentiality of the oblique cameras
for large-scale documentation purpose. The collection of
data have been carried out in several steps. The
measurement of the topographic network by using static
Figure 3: (left) The orthoimage from the UAV acquisition and
GPS/GNSS of 7 vertices. A group of Ground Control
(right) detail of top view of the vectorialized laser point cloud
Points (GCPs) were placed on the masonries using
paper targets, and then measured with a Total Station,
for the photogrammetric process and scans registration.

This work is licensed under a Creative Commons 4.0 International License (CC BY-NC-ND 4.0)
EDITORIAL UNIVERSITAT POLITCNICA DE VALNCIA 351
UAV OBLIQUE DATA AND LASER SCANNING IN AN EXCAVATED AREA

three approaches applied on the case study, if distinctly


evaluated, are indeed different, and then their scale is
diverse in terms of quality of image, richness of
information, time and resources consumption (human,
software, hardware). Thus, their selection depends
mostly by the characteristics of the sites and the
available time for the fieldwork. Neverthless the use of
TLS has to be combined anyway with cameras
acquisition for high quality texturing of the model. Then
time processing and human involvement in close-range
acquisitions can be defined as very competitive factors
thanks to the more and more automated algorithms of
digital photogrammetry now available. It is importan to
underline that all these technology improvements in
Figure 4: The digitalization of section G-G from LiDAR data
digital photogrammetry softwares are largely affected by
and the 4 samples for analysis: (a) masonry wall; (b) pilaster computer hardware configuration. The geometrical
with niches, mixed stones-bricks walls; (c) central apse part, definition provided from the UAV photogrammetry
mixed stones-bricks walls; (d) lateral apse and walls ridges. process is radically different in terms of scale, but this
technique complete necessarily the terrestrial survey
and reach a good definition of DSM, thanks to the
combinated nadiral and oblique cameras, now
increasingly use in research applications.

Figure 5: Images of the DSM in colors range by elevation (left)


and the RGB textured one (right).

Since the aims of the 3D survey were the definition of


the whole volume and the determination of the main Figure 6: Example of Density analysis on Sample (a): point
geometries of the fountain, a comparison between clouds (left) from LiDAR, npt 2.617.678, Density 186.995
different results and employed techniques is needed. pt/m2; (centre) close-range phot., npt 2.734.377, Density
Firstly, for the production of the plan it was necessary to 205.860 pt/m2; (right) UAV npt 44.823, Density 3.489 pt/m2.
integrate the terrestrial LiDAR data and the orthoimages
from the aerial ones. The laser point cloud defined in a Table 1: Results from comparison between points clouds.
hight detailed way all the masonries features in order to Close-range and UAV photogrammetry were matched to LiDAR
cover different levels of the entire fountain area, both in by cloud/cloud distance tool in Cloud Compare software.
the caves and in the bassin. The coverage of the TLS Absolute distances (m) Lidar - close-r Lidar - UAV
was generally very good and wide but lacked in details in
the higher portions of the masonries. Otherwise, the data Sample (a) mean 0.012 0.016
from UAV, despite the intrinsic different scale, allowed st. dev. 0.007 0.013
filling deficiencies of the terrestrial point cloud and
generating a complete 3D model of the site. In Table1 Sample (b) mean 0.009 0.017
some comparison data on one (Fig. 6) of the samples st. dev. 0.011 0.021
tested for a global evaluation on the 3 different sensors.
Sample (c) mean 0.004 0.021

3. Conclusions st. dev. 0.003 0.030

In case these complex objects or sites, the integration of Sample (b) mean -- 0.014
multisensors data can be pinpointed, according to st. dev. -- 0.015
acheived results, as the best solution for a multiscale 3D
documentation. Infact, the level of detail offered by the

References
BALLETTI, C., GUERRA, F., SCOCCA, V. and GOTTARDI, C., 2015. 3D integrated methodologies for the
documentation and the virtual reconstruction of an archaeological site. The International Archives of
Photogrammetry, Remote Sensing and Spatial Information Sciences, 40(5), 215.
BOEHLER, W. and MARBS, A., 2004. 3D scanning and photogrammetry for heritage recording: a comparison.
Proceedings of the 12th International Conference on Geoinformatics, University of Gavle, Sweden, pp. 291298
CORNAGLIA, P., 1994. Giardini di marmo ritrovati. La geografia del gusto in un secolo di cantiere a Venaria Reale
(1699-1798). Torino: Lindau. 226 pages.

This work is licensed under a Creative Commons 4.0 International License (CC BY-NC-ND 4.0)
EDITORIAL UNIVERSITAT POLITCNICA DE VALNCIA 352
Chiabrando, Span, Sammartano, Teppati, 2016.

LANDESCHI, G., DELL'UNTO, N., LUNDQVIST, K., FERDANI, D., CAMPANARO, D.M. and LEANDER, A.M., 2016. 3D-
GIS as a platform for visual analysis: Investigating a Pompeian house, Journal of Archaeological Science, Vol. 65,
pp. 103-113
LERMA, J.L., SEGU, A.E., CABRELLES, M., HADDAD, N., NAVARRO, S. and AKASHEH, T., 2011. Integration of laser
scanning and imagery for photorealistic 3D architectural documentation. INTECH Open Access Publisher.
PATIAS, P., 2013. Overview of applications of close-range photogrammetry and vision techniques in Architecture and
Archaeology, in MCGLONE, C. (edited by), Manual of Photogrammetry, 6th edition, Asprs (American Society of
Photogrammetry and Remote Sensing), pp. 1093-1107.
RUPNIK, E., NEX, F., REMONDINO, F., 2014. Oblique multi-camera systems orientation and dense matching issues.
Int. Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences, Vol. 40 (3/W1), pp. 107-114
STRECHA, C., 2014. The rayClouda vision beyond the point cloud. FIG Congress 2014, Engaging the Challenges -
Enhancing the Relevance, Kuala Lumpur, Malaysia 16 21 June 2014
WIEDEMANN, A. and MORE, J., 2012. Orientation strategies for aerial oblique images. ISPRS Archives of
Photogrammetry, Remote Sens ing and Spatial Sciences, Vol. 39 (B1).
XIAO, J., GERKE, M., VOSSELMAN, G., 2012. Building extraction from oblique airborne imagery based on robust
faade detection, Int. Journal of Photogrammetry and Remote Sensing, 68, pp. 56 6

This work is licensed under a Creative Commons 4.0 International License (CC BY-NC-ND 4.0)
EDITORIAL UNIVERSITAT POLITCNICA DE VALNCIA 353

View publication stats

You might also like