Detecting Facade Damage on Moderate Damaged Type From High-Resolution Oblique Aerial Images

Abstract

Oblique aerial images provide a more comprehensive view for the geometric and texture information of both rooftop and facade of buildings, hence it is possible to precisely detect damage grading of building for a detailed and overall damage assessment after a disaster event. The detection of damaged to building facades can improve the accuracy of damage-type classification to support reconstruction after disaster events, especially in the case of moderate damaged buildings. In this paper, a novel approach for automatic detection of damaged facade based on local symmetry feature and the Gini Index using oblique aerial images is presented. First, facade is extracted from oblique images using three-dimensional texture mapping. Then, local symmetry points are detected in a sliding window, and we obtain the histogram bins of local symmetry points from vertical and horizontal direction. Finally, damaged and nondamaged of building facade are distinguished using Gini Index. An evaluation of experimental results, for a selected study site of the Beichuan earthquake ruins, Sichuan, China, show that this method is feasible and effective for detection of damaged facade.