Object Tracking Using Probabilistic Principal Component Analysis Based on Particle Filtering Framework
In this paper, an object tracking approach is introduced for color video sequences.The approach presents the integration of color distributions and probabilistic principal component analysis (PPCA) into particle filtering framework.Color distributions are robust to partial occlusion, are rotation and scale invariant and are calculated efficiently.Principal Component Analysis (PCA) is used to update the eigenbasis and the mean, which can reflect the appearance changes of the tracked object. And a low dimensional subspace representation of PPCA efficiently adapts to these changes of appearance of the target object.At the same time, a forgetting factor is incorporated into the updating process, which can be used to economize on processing time and enhance the efficiency of object tracking.Computer simulation experiments demonstrate the effectiveness and the robustness of the proposed tracking algorithm when the target object undergoes pose and scale changes, defilade and complex background.
object tracking probabilistic principal component analysis particle filter defilade complex background
Zhi-yan Xiang Tie-yong Cao Peng Zhang Tao Zhu Jing-feng Pan
Institute of Communications Engineering,PLA Univ.of Sci.& Tech.,Nanjing,210007,China Institute of Command Automation,PLA Univ.of Sci.& Tech.,Nanjing,210007,China
国际会议
厦门
英文
790-797
2011-07-08（万方平台首次上网日期，不代表论文的发表时间）