In order to improve the accuracy of target tracking

a target tracking algorithm based on twodimensional principal component analysis（ 2 DPCA） and sparse representation is proposed to deal with the changes of illumination

occlusion and posture during target tracking. In the Bayesian framework

2 DPCA and L2 normalization are used to present fast and robust target tracking algorithms. A new method of likelihood function representation is proposed

and the incremental subspace learning method is used to update redundancy dictionary to effectively suppress the tracking drift and to deal with the target occlusion problem. Through qualitative and quantitative analysis on challenging tracking video

experimental results show that this method is superior to the traditional tracking method in tracking precision.