基于背景抑制的4种红外小目标检测算法比较  

The Comparison of Four Detection Performances for Small Targets in Infrared Images Based on Background Rejection

作　　者：余小英[1,2] 李凡生[1,2] 邵晓鹏[2] 

机构地区：[1]南宁师范高等专科学校物理与信息技术系,广西龙州532400 [2]西安电子科技大学技术物理学院,陕西西安710071

出　　处：《红外技术》2009年第5期287-290,294,共5页Infrared Technology

基　　金：陕西省自然科学基金项目(编号:2006F20)

摘　　要：背景抑制是红外小目标检测的关键技术。对空域高通滤波、最大中值滤波、单一结构元素形态学和多结构元素形态学的背景抑制原理进行分析,并在云杂波背景条件下对4种算法进行仿真并作比较。仿真结果表明多结构元素形态学算法性能最好,最大中值滤波算法次之,单一结构元素形态学算法稍差些,高通滤波算法最差。The key technology in infrared small target detection is background Suppression. In the paper, the principle of the background suppression about the high-pass filter, the max-median filter, the single-structure element morphology and the multi-structural element morphology are analyzed. Further more, the simulation and comparison of four algorithms are conducted in the cloud-clutter background. The simulation results show that the performance of the multi-structural element morphology is the best, the max-median filter the second, the single-structural element morphology the third, and the high-pass filter the worst.

关 键 词：云杂波背景 背景抑制 小目标检测 多结构元素 

分 类 号：TN911.73[电子电信—通信与信息系统]