Fusion Segmentation Algorithm for SAR Images Based on HMT in Contourlet Domain and D-S Theory of Evidence

摘要

Utilizing the Contourlet's advantages of multiscale, localization, directionality and anisotropy, a new SAR image segmentation algorithm based on hidden Markov tree (HMT) in Contourlet domain and dempster-shafer (D-S) theory of evidence is proposed in this paper. The algorithm extends the hidden Markov tree framework to Contourlet domain and fuses the clustering and persistence of Contourlet transform using HMT model and D-S theory, and then, we deduce the maximum a posterior (MAP) segmentation equation for the new fusion model. The algorithm is used to segment the real SAR images. Experimental results and analysis show that the proposed algorithm effectively reduces the influence of multiplicative speckle noise, improves the segmentation accuracy and provides a better visual quality for SAR images over the algorithms based on HMT-MRF in the wavelet domain, HMT and MRF in the Contourlet domain, respectively.