关键词: 运动目标跟踪, 二维主成分分析, 稀疏表示, 粒子滤波算法

Abstract: In order to improve the target tracking accuracy, for the tracking algorithm’s illumination, occlusion and pose variation problem of poor robustness, tracking algorithm is presented for two-dimensional principal component analysis and sparse representation of the target. The sparse representation of the two-dimensional principal component is analized to reduce the dimension of data, and to reduce the computational complexity; the particle filter algorithm is used to track moving target in image sequences; the algorithm is tested in simulation experiment for its performance. The simulation results show that, compared with other video target tracking algorithm, this algorithm can track moving target more reliably than other algorithm in image sequences, and has good robustness and the appearance from changes caused by the process of target tracking of illumination and pose variation. The algorithm has obvious advantages in more serious occlusion target tracking case.

Key words: object tracking, two dimensional principal component analysis, sparse representation, particle filter algorithm