针对UWBSAR图像中的非均匀背景目标检测问题,提出了一种CFAR检测方法.该方法运用变量索引和秩序统计量CFAR检测,可以在杂波背景的非均匀区域,自动根据检测单元的位置选择合适的估计杂波功率的样本,能使检测器在多目标环境和不规则非均匀背景交界处保持CFAR性能.试验结果表明,该方法在UWBSAR非均匀背景中的目标检测性能,优于传统的单元平均CFAR检测算法.

A CFAR target detection method is proposed for nonhomogeneous environment in UWBSAR image with variability index and order-statistics CFAR detection. The method can select appropriate cells in non-homogenous regions for clutter environment power estimation. The designed detector can keep CFAR performance in multi-targets environment and irregular boundary of non-homogenous regions. Experimental results show that the novel method can offer better detection performance than traditional cell average CFAR detection methods.