Urban change detection using coherence and intensity characteristics of multi-temporal SAR imagery

Change detection using multi-temporal SAR images is one of the most important applications of remote sensing technology. A new method for unsupervised change detection in urban area with multi-temporal SAR images is proposed in this paper. The method operates in two steps: change measures computation and unsupervised 2-D thresholding. In the first step, two change measures, backscattering intensity variation and long-term coherence variation, are defined. In the second step, an unsupervised 2-D thresholding technique based on maximum 2-D Fuzzy entropy criterion is introduced, which is performed on the two difference images derived from the two change measures to produce an accurate change-detection map with two classes ¿change¿ and ¿non-change¿. The parameters of Fuzzy entropy are optimized by genetic algorithm (GA). The effectiveness of the new method is validated by four ERS SAR images of Nanjing area acquired in 1997 and 1999.