Abstract: Due to the inconsistency of multitemporal images’ boundaries and spatial correspondence in the task of object based synthetic aperture radar （SAR） image change detection, a superpixel cosegmentation for SAR image change detection is proposed. Firstly, the pixel intensity similarities between the two pixels of the multitemporal SAR images are calculated respectively, which are then combined using a weight factor to form a new similarity measurement. Additionally, the edge magnitudes of the two multitemporal SAR images as well as their log ratio image are detected, and the maximum value among which is chosen to form a binary edge map image. Finally, the weighted similarity based on pixel intensity, location distance and edge information is used to replace the CIELAB space similarity for  local clustering in simple linear iterative clustering. The multitemporal SAR images are then cosegmented with accurate boundaries and spatial correspondence. The experimental  results conducted on a pair of simulated SAR images and a pair of real world multitemporal SAR images demonstrate that the boundary recall, undersegmentation error and achievable segmentation accuracy of the proposed method are better than those of other four state of the art methods.

Key words: synthetic aperture radar, superpixel segmentation, image cosegmentation, change detection