基于光流场分割的伪装色运动目标检测  

Camouflage Color Moving Object Detection Based on Optical Flow Field Segmentation

作　　者：周静 窦一民[1,2,3] 李金屏 ZHOU Jing;DOU Yimin;LI Jinping(School of Information Science and Engineering,University of Jinan,Jinan 250022,Shandong,China;Shandong Provincial Key Laboratory of Network Based Intelligent Computing,University of Jinan,Jinan 250022,Shandong,China;Shandong College and University Key Laboratory of Information Processing and Cognitive Computing in 13th Five-year,University of Jinan,Jinan 250022,Shandong,China)

机构地区：[1]济南大学信息科学与工程学院,山东济南250022 [2]济南大学山东省网络环境智能计算技术重点实验室,山东济南250022 [3]济南大学山东省“十三五”高校信息处理与认知计算重点实验室,山东济南250022

出　　处：《济南大学学报（自然科学版）》2020年第4期328-334,共7页Journal of University of Jinan(Science and Technology)

基　　金：国家自然科学基金项目(61701192);山东省重点研发计划项目(2017CXGC0810);山东省教育科学规划“教育招生考试科学研究专设课题”(ZK1337212B008)。

摘　　要：为了解决伪装色运动目标难以检测的问题,通过分析动态背景下具有伪装色运动目标的运动变化规律,提出一种基于光流场分割的运动目标检测方法;在具有伪装活动目标且背景动态变化的视频中采用稠密光流算法计算出视频序列的运动光流场,根据运动目标与背景运动方式的差别对运动光流场进行分析,利用K-均值算法完成运动目标与背景的分割,获取运动目标区域。实验结果表明,在摄像头固定情况下,该方法能在变化的背景中有效检测出伪装色运动目标。To solve the detection problem of the camouflage color moving object,a new method based on optical flow field segmentation was proposed by analyzing the moving rules of camouflage moving object in dynamic background.Dense optical flow algorithm was used to calculate the moving optical flow field of video sequence with the camouflage object and the background dynamic change.Then,the moving optical flow field was analyzed according to the difference between the motion mode of the moving object and the background.K-means algorithm was employed to complete the segmentation of the moving objects in the background.The experimental results show that,when the camera is fixed,the method can effectively detect the camouflaged moving objects in the dynamic background.

关 键 词：伪装色 动态背景 运动目标检测 光流场 K-均值算法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]