Change Detection in Remote Sensing SAR Image Using a Ratio-Based Operator

Change detection in Synthetic Aperture Radar (SAR) image is a critical task as the images contain significant speckle noise. Moreover, the intensity differences between the input SAR images is another challenging factor in change detection. In order to address these issues, a novel change detection method is proposed by using ratio operator which is very effective to handle speckle noise. At first, the input images are filtered through an infinite symmetric exponential filter (ISEF). After that ratio operation is performed between each sample point of images with its neighbourhoods. Then, this ratio values are used to create binary feature vectors for the input images. The binary feature vectors are compared in the corresponding positions of the images to detect the changes. Experiments of different sets of SAR images demonstrate the effectiveness of the developed change detection method.