2009 Volume 9 Issue 4 Pages 4_26-4_38
In a large-scale earthquake, it is important to understand road conditions over a wide area to establish vehicular and evacuation routes. This will allow us to carry out rescue and support activities rapidly. We propose a method of detecting road-blockage areas in mountainous districts. We use satellite images taken before the disaster and aerial images taken after the disaster. Since we can rapidly obtain information on a wide area from aerial images, we can use the images effectively to understand road conditions. We detect damaged areas by comparing the two images and using the digital elevation model (DEM). Then, we detect road-blockage areas using connectivity relationships between the damaged areas. Finally, we project the results obtained onto the digital map. We developed relevant software on the computer, and we analyzed actual images to evaluate the effectiveness of our method.
