CMRT09

Stilla, Uwe

Bibliographic data

Monograph

Persistent identifier:: 856955019

Author:: Stilla, Uwe

Title:: CMRT09

Sub title:: object extraction for 3D city models, road databases, and traffic monitoring ; concepts, algorithms and evaluation ; Paris, France, September 3 - 4, 2009 ; [joint conference of ISPRS working groups III/4 and III/5]

Scope:: X, 234 Seiten

Year of publication:: 2009

Place of publication:: Lemmer

Publisher of the original:: GITC

Identifier (digital):: 856955019

Illustration:: Illustrationen, Diagramme, Karten

Language:: English

Usage licence:: Attribution 4.0 International (CC BY 4.0)

Publisher of the digital copy:: Technische Informationsbibliothek Hannover

Place of publication of the digital copy:: Hannover

Year of publication of the original:: 2016

Document type:: Monograph

Collection:: Earth sciences

Chapter

Title:: A TEST OF AUTOMATIC BUILDING CHANGE DETECTION APPROACHES Nicolas Champion, Franz Rottensteiner, Leena Matikainen, Xinlian Liang, Juha Hyyppä and Brian P. Olsen

Document type:: Monograph

Structure type:: Chapter

CMRT09: Object Extraction for 3D City Models, Road Databases and Traffic Monitoring - Concepts, Algorithms, and Evaluation (a) Matikainen et al., 2007 (b) Champion, 2007 (e) Rottensteiner, 2008 (f) Champion, 2007 (g) Rottensteiner, 2008 (h) Olsen and Knudsen, 2005 Figure 3. Evaluation Details (same colour code as Figure 1). FP new cases related to DSM errors (shadow areas), in Marseille streets (a)-(b) and Toulouse (c)-(d); (e)-(f) FN new cases (small changes); (g)-(h) FP new buildings related to bridges. completeness rates for demolished buildings and the high correctness for unchanged buildings that could be achieved in these contexts highlight the effectiveness of the presented approaches in verifying the existing objects in the databases. The main limitation in terms of qualitative efficiency concerns the relatively high number of FN new buildings - up to 12.1% in the Marseille test area with (Rottensteiner, 2008) - that are mostly related to the object change size. The economical efficiency of the presented approaches seems to be promising, with 80-90% of the existing buildings requiring no further attention by the operator. These buildings are reported to be unchanged, which saves a considerable amount of manual work. In terms of the economical efficiency, the main limitation is a high number of FP demolished buildings that have to be inspected unnecessarily. Again, this is mainly caused by problems in detecting small changes. Areas of improvement should concern input data and methodologies. Thus, the resolution of LIDAR data (1 point/m 2 ) used in this test appeared to be critical for the change detection performance: using higher density LIDAR data (e.g. 5-10 points / m 2 ) should improve the situation. As far as methodology is concerned, new primitives should be used in the algorithms, in particular 3D primitives (representing e.g. the 3D roof planes or building outlines) that can now be reliably reconstructed with the 3D acquisition capabilities, offered by recent airbome/spacebome sensors. Another concern should be the improvement of the scene models used in object detection such that they can deal with different object classes and their mutual interactions. By incorporating different object classes and considering context in the extraction process, several object classes could be detected simultaneously, and the extraction accuracy of all interacting objects could be improved. In this project, we learned how difficult it is to compare approaches of very different designs. To carry out a fair test, we chose to use the building label images and to limit the type of changes to demolished and new buildings. In addition, we chose to compare the building label images to the initial vector database, basing on a coverage rate featured by the parameter T h . Further investigations are necessary to study the actual impact of this parameter on the completeness and correctness rates. However, if we are aware of these drawbacks, we think that this scheme was sufficient to bring out some interesting findings. We also hope that our results - in conjunction with those of e.g. the ARMURS 3 project - will be helpful to create a nucleus of interested people, both in academia and private sector, and to speed up the progress in the vector change detection field. REFERENCES Breiman, L., Friedman, J. H., Olshen, R. A., Stone, C. J., 1984. Classification and regression trees. The Wadsworth Statistics / Probability Series, Wadsworth, Inc., Belmont, California. Champion, N., 2007. 2D building change detection from high resolution aerial images and correlation Digital Surface Models. In: IAPRSIS XXXVI-3/W49A, pp. 197-202. Champion, N., Boldo, D., 2006. A robust algorithm for estimating Digital Terrain Models from Digital Surface Models in dense urban areas. In: IAPRSIS XXXVI-3, pp. 111-116. N. Champion, L. Matikainen, F. Rottensteiner, X. Liang, J. Hyyppa, 2008. A test of 2D building change detection methods: Comparison, evaluation and perspectives. In: IAPRSIS XXXVII - B4, pp. 297-304. Heipke, C., Mayer, H., Wiedemann, C., Jamet, O., 1997. Automated reconstruction of topographic objects from aerial images using vectorized map information. In: IAPRS, XXX11, pp. 47-56. Kumar, S. and Hebert, M.„ 2006. Discriminative random fields. International Journal of Computer Vision 68(2), pp. 179-201. Matikainen, L., Kaartinen, K., Hyyppa, J., 2007. Classification tree based building detection from laser scanner and aerial image data. In: IAPRSIS XXXVI, pp. 280-287. Mayer, H., 2008. Object extraction in photogrammetric computer vision. ISPRS Journal of Photogrammetry and Remote Sensing 63(2008), pp. 213-222. Olsen, B., Knudsen, T., 2005. Automated change detection for validation and update of geodata. In: Proceedings of 6th Geomatic Week, Barcelona, Spain. Pierrot-Deseilligny, M., Paparoditis, N., 2006. An optimization-based surface reconstruction from Spot5- HRS stereo imagery. In: IAPRSIS XXXVI-1/W41, pp. 73-77. Rottensteiner, F., 2008 Automated updating of building data bases from digital surface models and multi-spectral images. In: IAPRSIS XXXVII - B3A, pp. 265-270. 3 http://www.armurs.ulb.ac.be. Last visited: 30 June 2009. 150

Access restriction

Copyright

fullscreen: CMRT09

Monograph

Chapter

Table of contents

Cite and reuse

Monograph

Chapter

Image

Image fragment

Citation links

Citation recommendation