基于方位散射熵的建筑物特征提取方法

Building Feature Extraction Method Based on Azimuth Scattering Entropy

在超宽带穿墙雷达成像领域,交叉极化雷达能有效识别建筑物角散射中心,而交叉极化接收到的回波信号较弱,成像中的耦合信号得到增强,角散射信号不易识别。对此提出一种基于方位散射熵的建筑物特征提取方法。该方法首先利用散射体交叉极化相关性对交叉极化成像结果进行加权提取角散射中心,然后通过方位散射熵滤除墙体杂波影响、增强墙角散射幅度,最后使用循环迭代的中心定位算法得到精确的墙角散射中心。仿真和实验数据结果表明,该方法通过角散射体的极化特性和方位角属性可以准确地提取建筑物角散射中心。

In the field of UWB radar imaging,cross polarization radar can effectively identify the scattering center of building corner,but weak the echo signal received in cross polarization.The coupling signal in the image is enhanced and the corner scattering signal is not easy to recognize.In this paper,a building feature extraction method based on azimuth scattering entropy is proposed.This method uses the scattering cross polarization correlation to weight the cross polarization imaging results and extract the scattering centers,and then filters out the influence of wall clutter and enhances the scattering amplitude of the corner through the azimuth scattering entropy.Finally,the center location algorithm finally using cyclic iteration is adopted to get accurate corner scattering center.The simulation and experimental results show that the scattering center of the building corner can be accurately extracted by making use of the polarization and azimuth properties of corner scatterer.