基于部分背景加权更新的均值漂移跟踪算法

Mean-Shift tracking algorithm based on background weighted update approach

为了解决均值漂移跟踪算法中背景对目标定位的扰动,提出了一种基于颜色和纹理混合特征以及采用背景加权更新的改进算法。改进算法先将原始视频序列RGB帧图像转换为HSV颜色空间表示,然后分别在H、S通道上提取颜色特征,在V通道上用LBP描述符提取纹理特征,在此基础上为目标区域和背景区域建立三维颜色纹理混合直方图作为其描述符;在对象的跟踪过程中,通过巴氏系数选择性地加权更新部分背景信息。实验结果表明,与基于全部背景更新策略相比,改进算法充分利用了颜色和纹理特征并加权更新背景信息,具有更高的可靠性和鲁棒性,具有更好的计算效率。

In order to solve the problems of background interference in the Mean-Shift tracking algorithm, this paper proposed an improved algorithm based on color and texture blending characteristics and background weighted update approach. The original RGB image was converted into the HSV color space, then color feature was extracted in the H, S channel and texture feature was extracted based on the LBP descriptor in the V channel. Base on this, this paper established the color-texture histogram of the object region and background. During object tracking, it updated the background region using weighted update approach according to the Bhattacharyya coefficient. The extensive experimental results show that, compared with the algorithm adopting the full background update approach with color or colortexture features, the improved algorithm makes full use of color and texture features and adopts weighted updated background region, and has a higher level of reliability and robustness and better execution efficiency.