New Method of Object Tracking under Complex Circumstance
Weiguo Zhang, Xin Tian · 2010
The original MeanShift algorithm uses color histogram as the target feature and does not take the objectives of spatial information into account. Therefore, there are some stability problems under the complex background. It leads to target losing easily. A new tracking algorithm named weighted mean of the centroids is proposed in the paper. In order to improve the robust of tracking process, the weighted mean of centroids corresponding to the color bins and the model update strategy is added. The problem of spatial information on color distribution is solved, and the situation of deviation from the real target caused by using single fixed target model is also avoided by using this algorithm. Furthermore, experimental results show that the improved tracking algorithm is better than the original meanshift algorithm.