计算机应用与软件2016，Vol.33Issue(11)：139-143,5.DOI:10.3969/j.issn.1000-386x.2016.11.033
连续同色调背景下运动目标自适应跟踪
ADAPTIVE MOVING TARGET TRACKING UNDER INCESSANT SAME HUE BACKGROUND
摘要
Abstract
Aiming at the problem that traditional Meanshift algorithm cannot track the target accurately and constantly under the interference of incessant same hue background,we proposed an improved Meanshift target tracking algorithm which is based on Kalman-optical flow (KOF). First,it models the feature points within the area of Meanshift window by hue space-based optical flow detection algorithm,and gets the coordinates of their images.Secondly,it uses the speed prediction of Kalman filter to exclude background feature points and obtains the space constraint condition which is based on the target model feature points.Finally,it combines the derived space constraint condition with the hue constraint condition of traditional Meanshift algorithm to construct new back projection histogram,and uses the new back projection graph as the probability density map of Meanshift for iteration,so as to complete the target tracking.Experimental results demonstrated that under the interference of incessant same hue background area this algorithm could still track targets accurately and constantly.关键词
Meanshift/目标跟踪/光流/Kalman滤波/同色调区域Key words
Meanshift/Target tracking/Optical flow/Kalman filter/Same hue area分类
信息技术与安全科学引用本文复制引用
陈从平,黄正,聂葳,石小涛..连续同色调背景下运动目标自适应跟踪[J].计算机应用与软件,2016,33(11):139-143,5.基金项目
国家自然科学基金项目（51475266）；三峡库区生态环境教育部工程研究中心开放基金项目（KF2015-11）；湖北省三峡大学培优基金项目（2015PY022）。 （）
