矩阵完备理论在雷达成像中的应用  

Application of matrix completion theory in radar imaging

作　　者：闫慧辰[1] 彭石宝[1] 张旭东[1] 

机构地区：[1]清华大学电子工程系,北京100084

出　　处：《清华大学学报（自然科学版）》2013年第4期546-549,共4页Journal of Tsinghua University(Science and Technology)

基　　金：中国博士后科学基金资助项目(2013M530626)

摘　　要：在合成孔径雷达(SAR)成像中,大量数据的采集和存储等过程制约了设备在大测绘带、高分辨率、长时间观测中的应用,同时也导致了能量的大量消耗。为减小数据量,该文将SAR成像技术同矩阵完备理论结合,利用成像场景形成的矩阵的低秩性,提出了新的合成孔径成像机制。该机制通过减少采样次数随机获得部分原始数据,继而形成场景的部分成像,而后利用矩阵完备理论对场景进行恢复。利用真实的SAR数据对新机制进行验证,得到了较好的成像结果。The imaging process of synthetic aperture radar （SAR） requires large amounts of data sampling and storage that restricts the use of SAR in wide swath, high resolution and long time applications. The sampling and storage also consumes much energy. This paper presents an algorithm that reduces the data sampling and storage that combines matrix completion theory with SAR imaging techniques. The low rank of the radar imaging scene matrix is exploited by the algorithm. An SAR imaging scheme is given to recover part of the imaging patch with fewer random samples. The patch image is then recovered using matrix completion theory. Tests with real SAR data give acceptable results.

关 键 词：雷达图像处理 矩阵完备 合成孔径成像 

分 类 号：TN957.52[电子电信—信号与信息处理]