Synthetic aperture radar image segmentation using edge entropy constrained Stochastic relaxation

A synthetic aperture radar (SAR) image segmentation method using the multi-level logistic (MLL) model and edge entropy constrained stochastic relaxation is proposed. Edge entropy is developed and combined with a stochastic relaxation process to get expected segmentation. Gamma distribution is used for SAR intensity data and MLL model for the underlying label image. Parameters of Gamma distribution are estimated using EM method. The proposed method is an iterative scheme consists of two alternating steps: to approximate the estimation of the pixel class labels and to estimate gamma distribution parameters. The weight of the prior part in goal energy function is increased slowly versus the increasing iteration times until the edge entropy value of segmentation reaches an experiential threshold. The segmentation results for synthetic and real SAR images show that the proposed method has a good performance.