基于卡尔曼滤波的多运动目标跟踪算法研究  

Research of Multiple Moving Objects Tracking Algorithm Based on Kalman Filter

作　　者：崔吉[1] 张燕超 

机构地区：[1]中国矿业大学徐海学院,徐州221000 [2]中石油天然气管道局第二工程分公司,徐州221008

出　　处：《影像技术》2010年第3期11-15,共5页Image Technology

摘　　要：针对多运动目标跟踪的实时性和鲁棒性问题,本文提出了一种基于卡尔曼滤波的多运动目标跟踪算法,该算法运用卡尔曼滤波预测目标的位置,并以目标的中心点坐标、面积和长宽比特征、一维HSV颜色直方图作为目标的特征对当前帧检测到的目标模板和预测区域内的目标进行匹配。实验证明,该算法可实时、稳定地跟踪复杂场景内的多运动目标,并能够解决目标遮挡问题。Aiming at the real time and robustness problems for multiple moving objects tracking,a kind of multiple moving objects tracking algorithm based on Kalman filter is proposed in this paper,the algorithm uses Kalman filter to predict the object position,and takes the coordinate of center,the area,the feature of length/width ratio,and the feature of one dimension HSV color histogram as the feature of the object,the similarity matching between the object templates and the predicted area.It is shown that by the experiments,the algorithm can track multiple moving objects under the complex background in real time and stably,and can solve the problem of multiple moving objects under the situation of occlusion well.

关 键 词：卡尔曼滤波 多运动目标 预测 跟踪 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]