基于Robinson双模板匹配及融合的点目标检测算法  

Point Detection Algorithm Based on Robinson Filter's Dual Patterns Matching and Merging

作　　者：陈喆[1] 顾国华[1] 

机构地区：[1]南京理工大学电子工程与光电技术学院,江苏南京210094

出　　处：《红外技术》2011年第6期340-344,共5页Infrared Technology

基　　金：部委预研究项目;编号:40405030103

摘　　要：为了提高复杂背景下红外预警系统对空中红外点目标的探测效果,对传统Robinson滤波器的模板应用进行了延伸,在背景抑制和背景分割的基础上,对单帧红外图像中点目标进行Robinson滤波器双模板匹配及融合检测。实验表明,与传统Robinson滤波相比,此方法既提高了点目标的检测概率,同时更好地抑制了背景,取得了较为理想的目标检测效果。In order to increase the infrared prewarning system＇s efficiency of detecting the infared point target in the air,the application of Robinson filter＇s patterns is extended in this article.The point target in the single infrared image is detected by matching and merging the dual Robinson filter＇s patterns,on the basis of background suppression and division.The experiments have proved that this method can increase the detection probability of the point target and suppress the background more obviously,and it can acquire a better efficiency of point target detection than traditional Robinson filter.

关 键 词：点目标检测 Robinson滤波 背景抑制 背景分割 双模板融合 

分 类 号：TN911[电子电信—通信与信息系统]