Change Detection in Synthetic Aperture Radar Images

In this paper, we introduce a new approach for detecting the changes in SAR images that is taken from same geographical areas but at different periods of time. In this particular change detecting strategy, firstly, the ratio operator is applied to the SAR images to get the difference image. Here, log ratio and gauss log ratio operator is used to get the difference image. Then the difference image is fused by DWT. Noise in the fused image is removed by NSCT, which will remove the noise while preserving the information, especially the edge information. This denoised image is converted to compressed vectors by compressed projection thus reducing the dimension. Finally the changed and unchanged region is detected by clustering using fuzzy c mean (FCM) clustering with a novel Markov Random Field (MRF) energy function which aids in modifying the membership of each pixel and thus improving the accuracy in change detection.