一种新的基于稀疏贝叶斯学习的ISAR成像方法  

A novel ISAR imaging method based on sparse Bayesian learning

作　　者：成萍 姜义成[1] 许荣庆[1] 

机构地区：[1]哈尔滨工业大学电子工程技术研究所,哈尔滨150001

出　　处：《哈尔滨工业大学学报》2007年第5期730-732,共3页Journal of Harbin Institute of Technology

摘　　要：将BP(基寻踪)算法和FOCUSS(Focal underdetermined systemsolver)算法用于ISAR成像中,由于它们具有一些缺点,在实际应用中受到很大的限制.但是SBL(稀疏贝叶斯)方法可以克服这些缺点.本文首次提出可以将SBL代替BP和FOCUSS用于ISAR成像中.真实ISAR数据的成像结果表明SBL是一种比BP和FOCUSS更有效,更有潜力的成像方法.When BP （Basis Pursuit） and FOCUSS （Focal underdetermined system solver） algorithms are applied in ISAR imaging, as they have some shortcomings, they are limited in application. But SBL （sparse Bayesian learning） method can overcome these shortcomings. In this paper, it is first proposed that SBL can be used in ISAR imaging replacing BP and FOCUSS. Imaging results of real data show SBL is a more effective and more promising ISAR imaging algorithm than BP and FOCUSS.

关 键 词：ISAR(逆合成孔径雷达) 稀疏信号表示 SBL(稀疏贝叶斯学习) BP(基寻踪) FOCUSS(Focal UNDERDETERMINED system solver) 

分 类 号：TN958[电子电信—信号与信息处理]