基于贝叶斯压缩感知的ISAR自聚焦成像  

An Autofocus Imaging Method for ISAR Based on Bayesian Compressive Sensing

作　　者：王天云[1,2] 陆新飞 孙麟[1] 陈畅[1] 陈卫东[1] 

机构地区：[1]中国科学技术大学电磁空间信息重点实验室,合肥230027 [2]中国卫星海上测控部江阴,214431

出　　处：《电子与信息学报》2015年第11期2719-2726,共8页Journal of Electronics & Information Technology

基　　金：国家自然科学基金(61172155,61401140);国家863计划(2013AA122903)~~

摘　　要：针对ISAR自聚焦成像,该文提出一种基于贝叶斯压缩感知的高分辨率成像算法。首先利用目标图像的稀疏特性构建级联形式的稀疏先验模型,同时将相位误差建模为均匀分布模型;然后基于最大后验准则,依据贝叶斯压缩感知理论交替迭代求解目标图像和相位误差。与传统稀疏方法相比,所提算法进一步利用了目标图像的联合稀疏信息,将ISAR CS成像转化为MMV联合稀疏优化问题的求解,可以有效改善自聚焦的精度以及成像质量。仿真结果验证了该算法的有效性。For Inverse Synthetic Aperture Radar （ISAR） autofocus imaging, this paper proposes a high-resolution imaging method based on Bayesian Compressed Sensing （BCS）. Firstly, according to the sparsity characteristics of target image, a sparse model with the hierarchical framework is established, which can achieve better approximation to the originall0norm. Then, the phase errors are assumed to obey the uniform distribution. Next, following thecriterion of Maximum A Posteriori （MAP）, target image and phase errors are solved using alternate iteration based on BCS theory. Compared with traditional methods, the proposed method further combines the joint sparse information of target image, and converts the ISAR CS imaging into solving a joint Multiple Measurement Vector （MMV） sparse optimization problem, which can improve both the autofocus precision and the imaging quality efficiently. Simulation results show the effectiveness of the proposed method.

关 键 词：逆合成孔径雷达 自聚焦技术 高分辨成像 贝叶斯压缩感知 

分 类 号：TN958[电子电信—信号与信息处理]