A New CFAR Ship Detection Algorithm Based on 2-D Joint Log-Normal Distribution in SAR Images

Abstract

The characteristic difference between targets and clutter is analyzed. Considering the ship target's gray intensity distribution and its shape difference compared to the clutter, in this letter, a new algorithm is presented based on correlation. The algorithm utilizes the strong gray intensity correlation in the ship target; also, the joint gray intensity distribution using 2-D joint log-normal distribution of a pixel with neighboring pixels in the clutter is modeled, which can be used for correlation-based joint constant false alarm rate detection. Using this algorithm, the false alarms caused by speckle and local background nonhomogeneity can be greatly reduced. The detection performance is much better.