Change detection based on region likelihood ratio in multitemporal SAR images

In this paper, a novel approach for change detection in multitemporal synthetic aperture radar (SAR) images is presented. The proposed approach based on region likelihood ratio feature detection exploits an edge fusion technique for the SAR images segmentation. Segmentation is the key process in change detection based on region feature and the proposed edge fusion technique of two segmentation images ensures the accuracy of the changed parts segmentation. The change-detection process is divided into three main steps: 1 ) two SAR images segmentation based on watershed algorithm; 2) edge fusion of two segmentation images; 3) the region likelihood ratio feature extraction for changed region detection on the fusion result. Experimental results obtained on real SAR images acquired by the ERS-1 confirm the effectiveness of the proposed approach