Small-scale change detection in urban areas with the high resolution SAR system PAMIR

We propose a new pixel-based method for computing the difference image in synthetic aperture radar change detection. It is shown that it is competitive with state-of-the art methods. Moreover, we review some global thresholding methods for image binarization, and applied them to small-scale change detection in urban areas using the high resolution SAR system PAMIR. Unfortunately, no segmentation method was found to be fully satisfactory.