基于局部结构张量的红外图像背景抑制  

Background Suppression for the Infrared Image Using the Local Structure Tensor

作　　者：胡惠玲[1] 秦翰林[1] 拜丽萍[1,2] 周慧鑫[1] 

机构地区：[1]西安电子科技大学技术物理学院,西安陕西710071 [2]中国电子科学研究院情报中心,北京100041

出　　处：《电子科技》2010年第3期4-6,共3页Electronic Science and Technology

基　　金：教育部科学技术研究重点基金资助项目(108114)

摘　　要：文中将局部结构张量与双边滤波相结合,提出了一种基于局部结构张量的红外图像结构化背景抑制算法。该算法采用局部结构张量的标准差作为双边滤波的执行判定,并利用双边滤波加权系数实现背景估计,从而实现结构化背景抑制和目标信息增强。实验结果显示,与二维最小均方滤波和偏微分方程算法相比,该算法对含有弱小目标的复杂背景具有良好抑制效果。A background suppression algorithm for the infrared image based on the local structure tensor is presented, which combines the local structure tensor and bilateral-filtering. In the algorithm, the standard derivative of the local structure tensor is used as the criterion for whether the bilateral-filtering is executed, and the weighed coefficient with bilateral-filtering is adopted to achieve background estimation. The structured background suppression and the target-signal enhancement are fulfilled. Experimental results show that the presented algorithm achieves better background suppression than the partial differential equation method and the two-dimensional LMS filter.

关 键 词：红外图像 背景抑制 局部结构张量 双边滤波 目标检测 

分 类 号：TN219[电子电信—物理电子学]