期刊名称：ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences
印刷版ISSN：2194-9042
电子版ISSN：2194-9050
出版年度：2006
卷号：XXXVI Part 3
出版社：Copernicus Publications
摘要：Laser scanning- or "LIDAR"-based systems show perfect performance during time-critical events, like data collection in disaster mana- gement. To this end the research reports from classification of damaged buildings, a special challenge, for which first results are given. Firstly, mathematical foundations compile some new insights which are specific for the mentioned task, like Bolzano's theorem and statistical tests based on an extended Gauss-Markov model. Secondly, it is presented how these tools support the segmentation of planar surfaces and the classification of TIN segments into undamaged and damaged elements as well as into connecting triangles. Results are presented for real data from a training area of the Swiss Military Disaster Relief. The present status of the investigation shows that clear assumptions can be made for damaged buildings. Further steps will fuse additional knowledge in terms of data and algorithms
