Multi-temporal SAR Image Change Detection Based on Distribution Divergence
基于分布差异的多时相SAR图像变化检测

According to the characteristic of SAR’s backscatter and imagery feature,the paper presents a distribution divergence method calculating the cross-entropy of Edgeworth polynomial to detect the change area in different temporal SAR images.The method simulates the SAR image’s probability distribution function with Edgeworth polynomial,and simulates the differentia index between two images with cross-entropy.The differentia image from differentia index reflects the images change,and reduces the influence of SAR images speckle noise.The detected results which is the different image segmentation results with CFAR demonstrate the method be of high detecting rate and low fault alarm rate.