Change Detection in SAR Images Based on FCM with Modified MRF Approach

摘要

This study proposes the change detection in syntactic aperture (SAR) images by using the Fuzzy C-Mean (FCM) clustering with the modified Markov Random Field (MRF) energy function. With the application of these methods the changed and unchanged regions can be identified. This approach works based on two important functions such as reducing noise in the image and modifying the membership functions for every pixels concerned. For reducing the effect of speckle noise and change detection, the proposed approach used to modify the objective function behalf of membership function where as the objective function is just returning the normal form of the FCM. By using like this, we will get the less consuming time that by recently improved FCM algorithm. The proposed method efficiently detecting the changes in the SAR image and the testing results proves that, the proposed method outperforms the existing implementations.