摘要
利用主动表观模型(AAM)可以对视频序列中人脸进行特征点定位,当目标对象与初始位置偏离过大时,就会使拟合过程陷入局部最小,使迭代无法收敛到正确位置,造成定位失败。针对此问题,提出了一种基于强跟踪滤波器(STF)预测的AAM(STF-AAM)人脸特征点跟踪方法。首先,将视频中头部运动看成动态系统,然后利用强跟踪滤波器对其进行预测跟踪,从而找到每一帧的拟合初始位置并进行拟合运算。由于视频序列中每一帧中的拟合初始位置都能被快速找到,从而取得了比较精确、快速的跟踪结果。实验结果表明,所提方法与传统方法相比在保证拟合精度的同时,提高了算法的跟踪定位速度。
Active Appearance Model （AAM） can locate facial feature points of video sequences. When the initial position is far away from the destination, the fitting process often falls into local minimum, so that the iteration cannot converge to the correct location, resulting in locating failure. Concerning this problem, a facial feature point tracking method of AAM using prediction of strong tracking filter （STF-AAM） was proposed. Firstly, it viewed the head movement in the video as a dynamic system and used Strong Tracking Filter （STF） to predict and track it. So the fitting initial position of each frame was found and fitting algorithm was executed. This method could find the fitting initial position of each frame of video sequences and achieve a more accurate and more rapid tracking result. The experimental results show that the proposed method performs better than the traditional method in the tracking speed along with the fitting accuracy.
出处
《计算机应用》
CSCD
北大核心
2013年第2期511-514,共4页
journal of Computer Applications
基金
国家自然科学基金资助项目(60962005
61261037)
关键词
主动表观模型
强跟踪滤波器
动态系统
拟合初始位置
特征点跟踪
Active Appearance Model （AAM）
Strong Tracking Filter （STF）
dynamic system
fitting initial position
feature point tracking
