一种基于自适应背景杂波模型的宽幅SAR图像CFAR舰船检测算法

New CFAR Ship Detection Algorithm based on Adaptive Background Clutter Model in Wide Swath SAR Images

利用海洋宽幅SAR图像进行大范围海域舰船检测在海洋监视、军事侦察等方面具有重要应用。由于海况的复杂性,宽幅SAR图像背景杂波特性随海域不同而变化。采用双参数CFAR检测算法和基于K分布CFAR检测算法在处理宽幅SAR图像时,由于在待检测的所有区域采用同种背景杂波模型,导致使用的杂波模型在不适应区域失配,使CFAR检测性能下降。针对这个问题,提出了一种基于自适应背景杂波模型的CFAR宽幅SAR图像舰船检测算法,该算法通过背景窗口的多尺度统计方差判断目标所处的杂波环境,自适应选择对应的背景杂波分布模型,最后根据已知的恒虚警率及选择的杂波概率密度函数进行CFAR检测。对20多幅宽幅SAR图像进行了试验,实验结果表明:该算法在检测精度上有明显的改善。

In recent year,ship detection of wide swath SAR images have been widely used in ocean surveil- lance and military reconnaissance. The background clutter property of wide swath SAR images ranges apply in different image regions due to the complex sea conditions. Two parameter CFAR detector and K-distri- bution-based CFAR detector use the same distribution model which estimates the background clutter to de- tect the whole area. The used model is not fit for some regions, making higher loss of CFAR, bringing down the test performance. In this paper,a novel CFAR ship detection algorithm is presented which chooses the background clutter distribution model according to the multi-scale statistical variance：that is, choosing log- normal distribution in a uniform region and K-distribution in a non-uniform region. Then a threshold of the constant false alarm rate and the probability density function can be derived by the CFAR detector. Experi- mental results from 20 different wide swath SAR images are given to demonstrate that the proposed algo- rithm decreases the false alarms effectively and has high practical value.