基于SIFT子图像融合的直角坐标FBP算法  

Based on SIFT sub-image fusion for rectangular SAR FBP algorithm

作　　者：张佳琪 李凉海 张振华 左绍山 ZHANG Jiaqi;LI Lianghai;ZHANG Zhenhua;ZUO Shaoshan(Beijing Research Institute of Telemetry,Beijing 100076,China)

机构地区：[1]北京遥测技术研究所,北京100076

出　　处：《遥测遥控》2019年第6期62-70,共9页Journal of Telemetry,Tracking and Command

基　　金：装备预研航天联合创新基金(天基THz视频SAR技术)。

摘　　要：针对合成孔径雷达实测数据成像算法计算数据量大的问题,提出一种基于SIFT子图像融合的直角坐标系下的FBP算法。算法首先在直角坐标系下对SAR子孔径信号进行后向投影成像得到对应的子图像,然后对各子图像进行相位梯度自聚焦实现相位补偿,最后采用基于尺度不变特性变换(SIFT)的SAR图像配准方法对聚焦后的子图像进行融合得到全孔径图像。算法计算量比传统极坐标系下的快速后向投影算法小,并且由于基于SIFT的融合是针对复图像的融合,因此融合后图像的分辨率较传统的融合方法有所提高。机载实测数据的处理结果验证了算法的有效性。A SAR FBP imaging algorithm based on the SIFT sub-aperture fusion is proposed aiming at the fault of large calculation quantity in the SAR traditional algorithm.First,the sub-aperture signal is imaged by backward projection in the system of rectangular coordinates to obtain the corresponding sub-image.Then the phase compensation of each sub-image is carried out by phase gradient autofocus algorithm.And then the registration method based on scale-invariant feature transformation is used to fuse the sub-image to the full aperture image.The imaging algorithm can reduce the computing time than traditional fast backward projection algorithm,and the fusion based on SIFT is for the fusion of complex images,so compared with the traditional fusion method,the resolution of the merged image can be improved.Finally,the imaged results of simulation and air-borne radar data both demonstrate the feasibility of the algorithm proposed in this paper.

关 键 词：子孔径成像 快速后向投影算法 SIFT SAR图像融合 

分 类 号：TN957.51[电子电信—信号与信息处理]