一种改进的基于时域的红外弱小目标检测算法  

An Improved Algorithm for Temporal Detection of Infrared Dim and Small Targets

作　　者：沈强[1] 陈少冲[1] 

机构地区：[1]西安电子科技大学技术物理学院,陕西西安710071

出　　处：《电子科技》2009年第3期1-3,6,共4页Electronic Science and Technology

基　　金：国家自然科学基金资助项目(60407012);陕西省自然科学基金资助项目(2006F20)

摘　　要：提出了一种基于时域的红外弱小目标检测算法。该方法根据目标与背景的时域特性,建立了不同像素类型的时域模型。首先,对不同像素模型的时域方差进行分析,滤除掉天空背景以及云内部的像素;然后,根据时域剖面线偏离其包络线的程度不同,对弱小目标进行检测。理论分析和实验结果表明文中算法对于低信噪比条件下的弱小目标检测,具有很好的检测性能。An improved algorithm for temporal detection of infrared dim and small targets is presented. The temporal models for different pixel type are given according to temporal characteristics of the target and background. First, the temporal variance for different pixel models is studied and clean sky and cloud are re- moved, then the dim and small target is detected according to the discrepancy of the temporal profile and its envelope. The theoretical analysis and the experimental result prove the validity of the detection algorithm for small targets in low SNR.

关 键 词：物理电子学 时域模型 弱小目标 时域剖面线 目标检测 

分 类 号：TN219[电子电信—物理电子学] TP391.41[自动化与计算机技术—计算机应用技术]