An Efficient Approach for Change Detection in SAR Images

This paper presents an unsupervised distribution free change detection for Synthetic Aperture Radar (SAR) images based on an image fusion strategy and novel fuzzy clustering algorithm with a Markov Random Field (MRF). By finding the mean ratio and log ratio on the two original images the SWT based fusion rules is applied for performing image fusion. This image fusion technique is introduced to generate the difference image. Log is used to find background image and foreground detected by mean ratio. A MRFFCM algorithm is proposed for classifying changed and unchanged regions in the fused difference image. It incorporates the information about spatial context in a novel fuzzy way for the purpose of enhancing the changed information and of reducing the effect of speckle noise. In order to reduce the effect of speckle noise in SAR image the MRF is established to modify the membership of each pixel. The proposed approach focuses on modifying the membership function instead of modifying the objective function to reduce the speckle noise. Its objective function returns to its original form of FCM which consumes less time than that of improved FCM algorithm. Then the approach modifies membership of each pixel according to a novel form of MRF energy function in which neighborhood pixel and their relationship are concerned. Experiments on real SAR images show that the image fusion strategy integrates the advantages of the log-ratio operator and the mean-ratio operator and gains a better performance and the proposed approach detect the real changes as well as remove the speckle noises.