A Precise and Stable Segmentation Algorithm of SAR Images Based on Random Weighting Method and Modified Level Set

Level set methods have been widely used for image segmentation due to their good boundary detection accuracy. In the context of synthetic aperture radar (SAR) image segmentation, the presence of speckles and the distribution estimation of SAR image remain important issues that may hinder the accuracy of any segmentation method based on level set methods. In this paper, we propose a more accurate and a stable segmentation method based on the random weighting method and modified threshold-based level set energy functional. The proposed method uses a level set evolution that is based on the minimization of an objective energy functional, whose propagation function is based on the
${\mathcal{ G}}^{0}$
statistical model, whose parameters are estimated by random weighting estimators, and the estimator is not affected by the hypothesized model and sampling number. In addition, a new regularization item and length term, which maintain the regularity of the level set function and contour respectively, were employed. The experimental results demonstrate that the proposed methodology has a good and stable capability of segmentation, both in homogeneous, heterogeneous, and extremely heterogeneous regions in SAR image.