Automated damage detection of buildings from aerial television images of the 2001 Gujarat, India earthquake

A severe earthquake, which heavily affected Gujarat Province in western India, occurred on 26 January 2001. Using aerial television images taken with a digital video camera from a light plane in the reconnaissance survey carried out by one of the authors, automated detection of areas with building damage, based on both the multi level slice method and the maximum likelihood classifier, was attempted. Areas of building damage were defined based on six image characteristics derived from color indices and edge elements. The resulting estimated distributions of areas of building damage were in good agreement with the actual situation, although we found that it was difficult to use the same threshold values and training data for all the images because of differences in some factors, such as the influence of sunshine and built environment in the two methods.

[1]  Masashi Matsuoka,et al.  Automated Detection of Damaged Buildings due to Earthquakes using Aerial HDTV and Photographs , 2001 .

[2]  Fumio Yamazaki,et al.  Automated detection of damaged buildings using aerial HDTV images , 2000, IGARSS 2000. IEEE 2000 International Geoscience and Remote Sensing Symposium. Taking the Pulse of the Planet: The Role of Remote Sensing in Managing the Environment. Proceedings (Cat. No.00CH37120).

[3]  Hajime Mitomi,et al.  Development of automated extraction method for building damage area based on maximum likelihood classifier , 2001 .