An Improved Automatic Ship Detection Method in SAR Images

This paper provides an improved automatic ship detection algorithm, which uses two-parameter CFAR algorithm based on Gauss-distribution to process the homogeneous imaging local area, and uses two-parameter CFAR algorithm based on K-distribution to process the heterogeneous imaging local area again. This improved algorithm keeps both the ability of traditional two-parameter CFAR algorithm´ good features, such as small computation quantity, easy to implement and so on, and the detection accuracy in complex sea conditions at the same time.