一种红外视频舰船目标检测跟踪方法  

A Method of Ship Target Detection and Tracking of Infrared Video

作　　者：孙艳丽 张鹏飞[2] 李晓龙[2] 孙炜玮 SUN Yanli;ZHANG Pengfei;LI Xiaolong;SUN Weiwei(Naval Aviation University,Yantai 264001;No.92975 Troops of PLA,Ningbo 315000)

机构地区：[1]海军航空大学,烟台264001 [2]中国人民解放军92975部队,宁波315000

出　　处：《计算机与数字工程》2024年第6期1680-1684,1814,共6页Computer & Digital Engineering

摘　　要：针对帧间差分算法中对于红外舰船目标特征提取不充分的问题,提出了一种改进的红外舰船目标检测算法。将两帧图像进行帧间差分后通过Ostu进行阈值分割,获得二值图像进行形态学处理,计算图像质心坐标,并在原图像上的质心坐标周围区域进行遍历搜索,获得平均灰度值后进一步进行阈值分割。通过对海天背景及复杂背景下的舰船目标跟踪进行实验分析,结果表明改进后的算法效果较好,能够将所关注的运动舰船目标检测出来并跟踪。Aiming at the problem of insufficient feature extraction of infrared ship target in inter-frame difference algorithm,an improved infrared ship target detection algorithm is proposed.After inter-frame difference,threshold segmentation is carried out through Ostu to obtain binary images for morphological processing,image centroid coordinates are calculated,and traversal search in the area around the centroid coordinates on the original image is performed.After obtaining the average gray value,threshold seg-mentation is further carried out.Through the experimental analysis of ship target tracking in sea-sky background and complex back-ground,the results show that the improved algorithm is effective and can detect and track the moving ship target.

关 键 词：红外图像 舰船目标检测 GUI软件设计 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]