基于达芬奇平台的运动目标跟踪算法  

Moving Target Tracking Algorithm Based on Davinci Platform

作　　者：董云飞[1] 谢红松 罗斌[1,2] 汤进[1,2] 

机构地区：[1]安徽大学计算机科学与技术学院,安徽合肥230601 [2]安徽省工业图像处理与分析重点实验室,安徽合肥230039

出　　处：《计算机与现代化》2015年第10期45-49,共5页Computer and Modernization

基　　金：国家863计划基金资助项目(2014AA015104);国家自然科学基金资助项目(61472002);国家科技支撑计划基金资助项目(2012BAH95F01)

摘　　要：在运动目标跟踪过程中,考虑到达芬奇平台的数据处理能力,结合Vibe算法和Mean Shift算法的思想提出一种改进的运动目标跟踪算法,同时为了解决Vibe算法在目标由运动变为静止后遗留在前景图像中的鬼影难以消除的问题,对视频图像序列中的每个像素点建立前景计数器,根据前景计数器的值判断背景模型的更新速度。针对达芬奇平台对算法进行优化,把改进的运动目标跟踪算法移植到TMS320DM6467T达芬奇平台下进行处理,实验结果表明改进的算法具有良好的跟踪效果。In the process of moving target tracking,considering the data processing ability of Davinci platform,combining the Vibe algorithm with the mean shift algorithm,an improved moving target tracking algorithm is presented. At the same time,in order to solve the problem that it is difficult to eliminate the ghosts left in the foreground image after the moving target becomes stationary state in the Vibe algorithm,we set up foreground counters for each pixel in the video image sequence,and judge the renewal speed of background model according to the value of counter. And the algorithm is optimized based on Davinci platform. At last,the improved moving target tracking algorithm is transplanted into TMS320DM6467 T platform. Experiment results show that the improved algorithm is of good tracking performance to moving target.

关 键 词：Vibe算法 运动目标跟踪 TMS320DM6467T mean SHIFT 达芬奇平台 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]