基于多信号分类法SAR动目标检测研究

Research on SAR Moving Target Detection Based on Multiple Signal Classification

在运动目标检测问题的研究中,动目标的图像会因为存在径向速度而产生方位偏移,造成图像偏离真实位置。为了解决不同径向速度的运动目标在SAR图像中的成像偏移位置可能位于同一分辨单元而难以检测的难点,提出一种基于MU-SIC(多信号分类)算法的双通道SAR动目标检测新方法。方法采用高分辨率MUSIC谱估计算法代替传统傅里叶变换进行运动目标的检测及运动参数的估计,通过对动目标重新聚焦得到其偏移矫正后的图像,完成运动目标的定位。仿真结果表明,方法改善了成像偏移位置在多个动目标的检测能力,提高了参数估计的精确度。

The SAR image of moving targets appears misplaced in the cross-range because of the slant-range velocity. Moving targets which perhap locating in the same resolution cell can not be imaged to correct positions. To solve the difficult point, a two-channel detecting method based onmultiple signal classification （MUSIC） technique is proposed according to the superposition model of multiple sine signals after displaced phase centre antenna （DPCA）. The new algorithm replaces the Fourier transform with high-resolution MUSIC spectrum estimation technology. At last, images of the moving targets are achieved, and the ground moving targets are relocated. The simulation results validated that the method can improve the capability of detecting multiple moving targets which locate in the same res- olution cell and estimate motion parameters more precisely.