摘要
针对传统的人体跟踪算法中,过于依靠三维信息的连续性,很难依靠少量残缺特征完成运动跟踪的情况,提出一种加入了关联分析的基于残缺特征关联的三维跟踪算法研究。该算法提高在传统的三维特征点匹配后,加入关联量的计算,计算前后运动的深度信息之间的关联程度,完成在特征残缺、丢失等情况下的运动跟踪过程。实验结果表明,该算法能够对人体三维运动进行比较准确的三维运动跟踪,取得了较为理想的效果。
According to the traditional human body tracking algorithms,too rely on three dimensional continuity of information,it is difficult to rely on a few fragmentary features complete motion tracking.This paper puts forward a joined the correlation analysis based on incomplete characteristics associated three dimensional tracking algorithms.Improve the traditional three dimensional feature point matching,then add the correlation calculation of quantity of calculation and the depth of the sports information correlation between degree,complete in feature fragmentary,lost,motion tracking process.The experimental results show that this algorithm can 3 d motion to human body comparison accurate threedimensional motion tracking.Achieved ideal result.
出处
《科技通报》
北大核心
2013年第6期104-106,共3页
Bulletin of Science and Technology
关键词
运动跟踪
关系信息
残缺特征
motion tracking
relationship information
incomplete characteristics
