基于CS的SAR旋转微动目标检测方法研究

Study on the rotational micro-motion target detection based on compressive Sensing for SAR system

本文提出一种基于压缩感知理论(CS)实现旋转微动目标检测的方法。首先根据正侧视SAR系统空间几何模型对旋转点目标和旋翼三叶片目标进行成像;然后根据旋转微动目标的成像特征和形式,构造包含线形和圆形元素的超完备字典,再利用Hough变换域的稀疏性,用压缩感知理论寻找待检测图像中是否包含旋转目标所成像的几种形式来检测转动目标;最后通过补偿校正的方法对旋转微动目标进行再次判别,排除虚假目标。实验结果证明了所提方法的有效性。

A method of rotational micro-motion target detection based on compressed sensing is proposed. Firstly,according to the geometry of the strip-map SAR,the imagery of rotational micro-motion target is gotten. Then we construct the over-complete dictionary which consisted of lines and circles based on SAR imaging characteristics of targets with rotational micro-motion. Compressed sensing can be used to detect if there are some kinds of forms of rotational micro-motion target in images by exploiting sparseness in the Hough transform domain. Finally,using the method of compensating correction,we can confirm the detection results and eliminate the false targets. The results of experiments indicate the effectiveness of the proposed method.