Analysis of Object Detection Based Approaches for Synthetic Aperture Radar Images

Day and night capability of Synthetic Aperture Radar (SAR) make it unique as compared to all the other sensors. Detection of objects such as marine vessels using SAR images helps to improve maritime surveillance of any country. In this article, an analysis of three different approaches has been discussed. The first approach is based on point target detection and other two approaches are based on degree of polarization (DOP). A full polarized ALOS PALSAR 1 data has been utilized to evaluate the performance of the all three algorithms. Degree of polarization-based approach is showing $95 \%$ accuracy which is better as compared to other state of art approaches.