A clustering Based Color Model and Fast Algorithm for Object Tracking

The paper presents a clustering based color model and develops a fast algorithm for object tracking. The color model is built upon k-means clustering, by which the color space of the object can be partitioned adaptively and the histogram bins can be determined accordingly. In addition, in each bin the multi-channel gray level is modelled as Gaussian distribution. A metric based on Chernov distance is defined to measure similarity between the reference model and the candidate model. The integral images are proposed for computation of mean vector and covariance matrix of color images, through which the similarity metric can be evaluated very fast. Comparisons with the well-known mean shift algorithm demonstrate the validity of the model and performance of the proposed algorithm