Research On The Algorithm Of Ship Target Detection In SAR Images

The ship targets detection in SAR images has great significance for our country as we have a coastline stretching more than 18000 kms. In civilian areas, by detecting the traffic situation to the designated sea area and harbor, we may constantly monitor the illegal or wrecked ships. In military areas, we may constantly monitor all the targeted naval vessels, including warships, supply ships and submarines of the enemy. As a result, in the last few decades, the development of this skill attracts domestic and overseas attention. This paper mainly discusses the combination of the classical algorithm and feature analysis method of ship targets detection in SAR image.Firstly, this paper describes the theoretical basis for the ship targets detection in SAR image. The discussion arranges from the ship detection based on image processing and template to the influence of the system characteristics, the environment of the imaging areas and the characteristics of the ship on the detection performance.Then the classification of the distribution model of the sea clutter is calculated and summarized. At the same time, the parameter estimations of the statistical distribution model of the sea clutter of some kinds of SAR images are calculated. And the results of various model are observed after experiments with specific SAR images.Next, the basic principle of identifying the targeted naval vessels in SAR images dealing with CFAR detectors is summarized. Thus, the principle, the implementation method and the applicability of some classical CFAR detectors are described. Dealing with targets detection of the specific SAR images, t he comparison of the performance of different detectors is found out, and the final results are obtained accordingly.Finally, the ship detection algorithm based on feature analysis is introduced and combined with the traditional CFAR detectors in order to achieve a more accurate detection results. Furthermore, based on the experiment of SAR image of specific river channels, the SAR images with large size, complex and multiple targets are processed, thus the superiority of the algorithm in this paper is proved.