A new method for building collapse detection jointly using interferometric and polarimetric information

Earthquake is one of the most catastrophic disasters to affect mankind especially in urban areas. Reliable and fast collapsed building detection is vitally important to timely rescue and scientific decision making. A modified detection method for collapsed buildings is proposed in this paper by jointly using interferometric information and polarimetric information. The method was tested on two single-polarimetric ALOS PALSAR images and one fully polarimetric Radarsat-2 image from Yushu earthquake in Qinghai province of China occurred in 2010. After compared with manual interpretation results of very high resolution (VHR) optical image, collapsed building detection accuracy can be 72.77%. The accuracy is improved by about 30% compared with that of only coherence based method.