A Backscattering-Suppression-Based Variational Level-Set Method for Segmentation of SAR Oil Slick Images

Abstract

Robust and accurate segmentation of oil slick regions from synthetic aperture radar satellite images plays a fundamental role for detecting and monitoring of oil spills. However, uneven intensity, high noise, and blurry boundary, which always exist in oil spill images, make the automatic segmentation of such images very difficult. In this paper, a two-stage method is developed for the segmentation of oil spill images. The first stage of our method is to obtain the enhanced image by suppressing the backscattering from an oil spill image. Once the enhanced image is obtained, then in the second stage, a variational segmentation model is presented for dealing with the enhanced image. The data term of the energy functional is constructed for the enhanced image in a piecewise constant way. In addition, a Cahn–Hilliard-type regularization term is introduced into the energy functional. The variational model is numerically solved by alternating minimization. Numerical experiments on $\text{65}$ oil spill images from ENVISAT show that the proposed method can obtain an overall accuracy of $\text{94} \%$ for dark spot segmentation and create limited false alarms and outperforms the two representative state-of-the-art methods in terms of the efficiency and accuracy.