摘要
为了提高人体运动的跟踪精度,提出基于图像序列的人体运动跟踪算法。首先对当前人体跟踪算法的研究现状进行分析,指出粒子滤波算法进行人体运动跟踪的不足;然后对粒子滤波算法进行改进,增加了采样粒子多样化,提高非线性人体运动目标跟踪性能,加快人体运动跟踪速度;最后采用仿真实验对人体运动跟踪算法的性能进行测试。实验结果表明,相对于其他人体运动跟踪算法,该算法提高了人体运动跟踪的准确性,而且人体运动跟踪的时间减少,具有更好的稳定性。
A human motion tracking algorithm based on image sequence is proposed to improve the tracking accuracy of human motion tracking.The current research status of the human tracking algorithm is analyzed.The insufficiency of the particle filtering algorithm to human motion tracking is pointed out.The particle filtering algorithm is modified to improve the diversity of the sampling particles and tracking performance of the nonlinear human movement target,and quicken the tracking speed of the human motion.The performance of the human motion tracking algorithm was tested with simulation experiment.The experimental results show that,in comparison with other human motion tracking algorithms,the algorithm has improved the tracking accuracy of human motion,reduced the tracking time of human motion,and has better stability.
作者
石曼曼
李玲
SHI Manman;LI Ling(Leshan Normal University,Leshan 614000,China)
出处
《现代电子技术》
北大核心
2017年第9期25-28,共4页
Modern Electronics Technique
基金
四川省科技厅支撑计划项目(2014SZ0107)
关键词
图像序列
人体运动
跟踪算法
仿真测试
粒子滤波算法
image sequence
human motion
tracking algorithm
simulation test
particle filtering algorithm
