Robust object tracking using joint color-texture histogram



LBP operator is an effective tool to measure the spatial structure of local image texture. To reduce the computational cost and improve the robustness of target representation, we proposed a joint color and LBP texture based mean shift tracking algorithm in this paper.

A mask of the target is formed based on its five major uniform LBPriu2 8,1 texture patterns and then the target is represented by using its color and texture features within the mask.

The proposed target representation model effectively extracts the edges and corners, which are important and robust features, of the object while suppressing the smooth background features. Experimental results indicate that the proposed method performs much better than the original color based method with fewer iteration numbers, especially in tracking objects that have similar color appearance to the background.

Tracking results of football sequence 2 by the target representation models



ref :

Ning, Jifeng, Lei Zhang, David Zhang, and Chengke Wu. “Robust object tracking using joint color-texture histogram.” International Journal of Pattern Recognition and Artificial Intelligence 23, no. 07 (2009): 1245-1263.

Robust object tracking via active feature selection

CPLEX Code of PhD thesis (Multi Objective Robust Aggregate Production Planning in a Supply Chain under Uncertainty)


There are no reviews yet.

Be the first to review “Robust object tracking using joint color-texture histogram”

Your email address will not be published. Required fields are marked *