The study of road damage detection based on high-resolution SAR image

This paper presents a technique for the detection of damaged road in spaceborne synthetic radar (SAR) images. Roads in SAR image can be modeled as line structures, and are extracted from image by Duda detector, and the roads are accurately detected by removing line structures which aren´t road information. We use a change detection algorithm based on Edgeworth approach and Kullback-Leibler divergence to obtain change information of images. Finally, we combined information of road and change detection result for detecting damaged road sections. This technique is applied on RadarSat-2 images that have a resolution of about 3m. The experimental results show that our method can detect mainly damaged road sections.