基于图割的雷达微弱目标检测

Radar Weak Target Detection Based on Graph-cuts

低信噪比、强机动条件下的雷达目标检测是目前的研究难点和热点。提出了一种基于图割的雷达目标全局优化检测方法。首先对距离-多普勒信号进行空域积累,形成一定的目标图像序列;然后对背景和目标建模,在最大后验概率准则下,采用图割算法对目标图像序列进行全局优化检测,获取雷达目标图像序列;最后利用卡尔曼滤波器预测目标图像特征,完成目标模型的更新。仿真结果表明,在低信噪比、强机动条件下,能够有效地实现对雷达微弱目标的检测。

Now radar weak target detection is the focus of research in the condition of low SNR and strong maneuverability .A method of radar weak target detection based on graph cuts is proposed in this paper . Firstly, the range-Doppler image sequence is constructed through the spatial accumulation; Secondly , based on the target and background models , the target sequence is detected by the optimization algorithm of the graph cuts under the criterion of MAP;Finally, the parameters of target model are predicted by Karman filter to update the target model .The tests demonstrate the effectiveness of this algorithm on detecting the radar weak target in the condition of low SNR , especially for the strong maneuverability .