超宽带合成孔径雷达图像中二面角目标的检测

Detection of Dihedral Targets in UWB-SAR Images

目标的散射特性常依赖于雷达和目标间的姿态角,这种依赖性在大积累角的UWB-SAR中尤为明显。通过分析SAR图像的频谱特点,提出一种归一化标准偏差滤波方法。在该方法中,SAR的复数图像首先通过一组带通二维滤波器,然后将滤波器组的输出图像序列合成一幅信杂比得到提高的图像。对仿真数据处理的结果表明,该方法的性能优于曾提出的方向匹配滤波方法。

The target's scattering characteristic is often dependent on target-radar attitude angle, which is especially prominent in the wide processing-angle UWB-SAR. After the characteristics of 2-D frequency domain of SAR images are analyzed, a method named normalized standard deviation filtering is proposed. A 2-D band-pass filter bank is designed to filter the complex SAR image and the output images are processed to get an image in which the signal-clutter-ratio is enhanced. The simulation result has proved that it is more efficient than the method of direction matching filtering proposed before.