基于奇异值分解的红外弱小目标背景抑制.pdfVIP

2009 6 30 3 秦翰林 等: 基 奇异值分解的红外弱小目标背景抑制 1 1 1 1, 2 秦翰林 , 周慧鑫 , 刘上乾 , 杨廷梧 (1. , 710071; 2. , 710089) : 复杂背景的抑制是红外弱小目标检测技术的 一个难题为解决这 个问题, 提出了基 奇异值分解的背景抑制算法从矩阵的角度出发, 通过对原图像进行奇异值分解, 将包含弱小目 标信息的图像矩阵分解到 一系列奇异值和奇异值矢量对应的子空间中, 然后通过定义的偏差指数 所确定的有效的奇异值来重构图像, 从而达到背景抑制的目的与二维最小均方误差算法比较, 实 验结果显示, 该算法对红外弱小目标复杂背景从主观视觉和数值指标都具有良好抑制效果 : ; ; ; ; : T N91173 : A : 1001- 5868( 2009) 03- 0473- 04 SVD for Infrared Dim and Small Target Backgro nd S ppression 1 1 1 1, 2 QIN Han-lin , ZH OU Hu-i in , LIU Shang-qian , YANG T ing-w u ( 1. School of Technical Physics, Xidian University, Xian 710071, CHN; 2. Chinese Flight Test Establishment, Xian 710089, CHN) Abstract: Complicated backgr ound suppression is a difficult problem for infrared dim and small target detection technique. T he algorithm based on singular value decomposition is presented. T he algor ithm decomposes image w hich contains dim and small target information to a series of sub-spaces corresponding to singular value and singular value vector. Then, accor ding to bias inde , specific effective singular values ar e e tracted to reconstruct tar get image for background suppression. Compared w ith conventional Tw o-dimensional LMS algorithm , through subject inspection and value inde , several groups of e perimental results demonstrate that the presented algorithm can suppr ess complicated background in infr ared dim and small target image effectively. Key words: infrared image; target detection; backgr ound suppression; singular value decomposition; bias inde 0 20 ,