一种提高距离扩展目标检测性能的新方法

New method for improving the performance of range-spread target detection

针对传统二进制积累的硬阈值滤噪性能的不足,提出了一种基于局部模糊阈值的距离扩展目标检测新方法.新方法对积累器的阈值设置进行了改进,使用局部模糊阈值抑制高分辨雷达距离像的噪声散射单元并保持目标散射单元,通过距离维和脉冲维能量积累实现多脉冲积累来提高检测性能.与目前广泛使用的基于散射点分布密度的广义似然比检测方法相比,新方法不需要目标的任何先验信息,易于实现恒虚警检测.实验结果表明:新方法相对于基于散射点分布密度的广义似然比检测方法的增益提高约2 dB,相对于二进制积累的增益提高约4 dB.

Aimed at the shortage of the conventional hard threshold of the binary integrator, a new method based on the local fuzzy thresholding map is proposed. The new method utilizes the local fuzzy threshold in range dimension to suppress noise-only cells and preserve target＇s strong scattering cells, Then, the energy integration in the range-pulse domain realizes multi-pulse integration to improve the detection performance. Compared with the commonly-used spatial scattering density generalized likelihood ratio test （SSD-GLRT）, by the proposed method one needn＇t know any a priori information on targets and this method has the constant false alarm rate （CFAR） characteristic. Experimental results show that the gain by the proposed method is increased about 2 dB with respect to that by SSD-GLRT, and about 4 dB with respect to that by the binary integrator.