Abstract: A novel particle filter approach was presented in this paper for robust object tracking using a multi-region based discriminative modeling mechanism. Conventional particle trackers usually take the target region as a whole and may easily be confused by the similar pixels from non-object region. In contrast, the proposed method highlights the most effective information of foreground/background discrimination and divided the object into a number of patches. By comparing the properties of within and between classes of the patch and background, the most informative patches are selected. Moreover, based on these selected patches, the proposed algorithm commanded both the photometric and spatial information in a Bayesian manner for target modeling. This discriminative appearance model was then used in updating step of the particle filter to measure the object region. Experimental results on several challenging video sequences had verified that the proposed method was compared very robust and effective with the traditional particle filter in many complicated scenes.