A fuzzy complete SAR processing chain for ship detection and velocity estimation

Environmental monitoring by means of Remote Sensing methodologies is one of the most interesting and important application field of image processing techniques. In this field, a very important role is played by Synthetic Aperture Radar (SAR) images, that posses an high spatial resolution and that can be produced in all weather conditions. This paper presents a complete digital signal processing chain for the automatic detection of ships in SAR images and for the computation of an estimate of ship velocities. A particular attention has been devoted to two problems: the prefiltering problem which aims to reduce the speckle noise in the image, and the robustness problem that is related to the optimization of the true-target detection/rejection capability. The whole chain has been first simulated and tested on a sequential machine, namely a Micro VAX II, and then it has been ported on a parallel architecture based on a hyper-cube of IMS T800 transputers.