一种快速检测红外小目标的形态学方法  

A Morphological Method for Fast Detection of Infrared Small Targets

作　　者：刘颖斌 曾钰惠 秦建华[1] LIU Yingbin;ZENG Yuhui;QIN Jianhua(College of Mechanical and Control Engineering,Guilin University of Technology,Guilin 541006,China;School of Mathematics and Statistics,Guangxi Normal University,Guilin 541006,China)

机构地区：[1]桂林理工大学机械与控制工程学院,广西桂林541006 [2]广西师范大学数学与统计学院,广西桂林541006

出　　处：《电子信息对抗技术》2024年第4期32-40,共9页Electronic Information Warfare Technology

摘　　要：红外小目标检测是一项具有挑战性的任务,传统形态学算法在小目标检测中容易受到复杂背景噪声的影响,检测效果不佳,且结构元素大小的选择容易影响检测效果。为解决该问题,通过结合多重的形态学算子,设置自适应加权的形态学算子,提出一种高效的小目标检测算法,最后在真实红外图像上进行测试。实验结果表明,所提出的方法在定量和定性方面都优于现有技术。Infrared small target detection is a challenging task.Traditional morphological algorithms in small target detection are susceptible to complex background noise,resulting in poor detection performance,and the selection of structural element sizes can easily affect the detection performance.To solve this problem,an efficient small target detection algorithm is proposed by combining multiple morphological operators and setting an adaptive weighted morphological operator.Finally,it is tested on real infrared images.The experimental results show that the proposed method outperforms existing technologies in both quantitative and qualitative aspects.

关 键 词：红外小目标检测 形态学算子 结构元素 背景抑制 自适应加权 

分 类 号：TN976[电子电信—信号与信息处理]