Target Detection in High-Resolution SAR Images Based on Modified Active Contour Model

In this paper, we propose a target detection approach based on the modified active contour model (ACM) for high-resoultion synthetic aperture radar (SAR) images. To overcome the effect of speckle noise, the statistical dissimilarity between local patches is measured by the Bhattacharyya distance combined with Gamma distribution. With this basis, a modified region-scalable fitting (RSF) model is constructed. Then modified RSF model is incorporated into the global minimization active contour (GMAC) framework to achieve global minimization. The proposed method avoids the existence of local minima and deals with the speckle noise. The experimental results with simulated and real SAR data sets demonstrate the effectiveness and efficiency of the proposed method.