一种红外小目标的图像检测方法  

Infrared Weak-target Image Detection

作　　者：周铭[1] 许少辉[2] 

机构地区：[1]华北光电技术研究所,北京100015 [2]中国电子产品可靠性与环境试验研究所,广东广州510610

出　　处：《半导体光电》2004年第3期224-227,共4页Semiconductor Optoelectronics

摘　　要：　针对复杂背景条件下的红外图像目标检测问题,将空域滤波和时域滤波相结合,提出了基于数学形态学的目标检测算法。该算法先对原始图像序列作能量膨胀累积以增强目标信息和提高信噪比,然后采用基于数学形态学的Top Hat变换空域滤波处理技术消除在目标尺度范围外的背景、干扰和噪声,达到探测目标的目的。仿真结果验证了该算法是一种实时、有效,且易于实现的目标探测方法。On infrared image target detection in complex background, an algorithm for targets detection based on mathematical morphology is put forward by combining spatial filtering with temporal filtering. Firstly, the original sequence images that have been dilated are accumulated in order to enhance target information and improve SNR, then the background disturbances and noises are eliminated by using Top-Hat transform spatial filtering method based on mathematical morphology so that the targets will be able to be detected. Simulation results show that the algorithm is a real-time, effective and practical target-detecting method.

关 键 词：红外图像 目标检测 数学形态变换 

分 类 号：TN219[电子电信—物理电子学]