Abstract: A target tracking algorithm of multiple features fusion is proposed for the problem that the tracking algorithm of single feature describing targets in Mean Shift tracking algorithm is not robust enough. The proposed method uses the HSV color feature and the ICLBP texture feature to establish the probability density of the target model. The background area is determined according to the target area, and the feature fusion weights are set and updated based on the distinguishing measure of the target and the background under different features. The proposed algorithm uses the feature fusion weighted coefficient to establish a multi-feature description of the target model, and achieves the target tracking in the Mean Shift algorithm framework. The results show that the proposed algorithm has good robustness to background interference and partial occlusion. Compared with the traditional Mean Shift tracking algorithm, the tracking effect is improved and the robustness is better.