一种新的红外小目标检测算法及其应用  

A New Method for Infrared Small Targets Detection and Its Application

作　　者：罗军辉[1] 姬红兵[1] 刘靳[1] 

机构地区：[1]西安电子科技大学电子工程学院,陕西西安710071

出　　处：《半导体光电》2007年第2期290-293,共4页Semiconductor Optoelectronics

基　　金：国防预研基金资助项目

摘　　要：针对红外探测系统中单帧红外图像中的小目标检测问题,在分析红外场景模型的基础上,提出了一种基于自适应噪声平滑和Robinson Guard背景杂波抑制的空间滤波算法。该方法对于起伏背景下红外小目标的检测和识别有良好的性能,且具有目标信息损失小的优点。利用ADSP-TS201S高性能数字信号处理器实现了该算法,并成功应用到全向红外搜索跟踪系统中。实验结果表明,在低信噪比情况下,所提出的单帧红外小目标检测算法有很好的检测性能。For small targets detection in single frame infrared （IR） image, a spatial filter algorithm based on an adaptive smooth filter and the Robinson Guard spatial filter is proposed, with the analysis of the IR scene model. The method solves the problem how to detect the small targets in the undulant background and has the merit with no target information loss. The algorithm is realized by the digital processor ADSP-TS201 with high performance and successfully used in an all directional IRST system. The experiment shows that this method has high detection performance for small targets in single frame at low SNR.

关 键 词：小目标检测 自适应平滑滤波 ROBINSON Guard空间滤波 ADSP—TS201S 全向红外搜索跟踪系统 

分 类 号：TN219[电子电信—物理电子学]