计算机应用 ›› 2010, Vol. 30 ›› Issue (3): 650-652.

• 图形图像处理 • 上一篇    下一篇

联合多特征的自动CamShift跟踪算法

卢璇1,雷航1,郝宗波2   

  1. 1. 电子科技大学计算机科学与工程学院
    2. 电子科技大学 计算机科学与工程学院
  • 收稿日期:2009-09-14 修回日期:2009-11-17 发布日期:2010-03-14 出版日期:2010-03-01
  • 通讯作者: 卢璇

Automatic CamShift tracking algorithm based on multi-feature

  • Received:2009-09-14 Revised:2009-11-17 Online:2010-03-14 Published:2010-03-01

摘要: 针对CamShift跟踪算法仅采用颜色作特征,易发生跟踪错误等问题,提出了一种基于特征融合的算法。采用改进的背景差分法自动检测目标,目标模型联合了颜色和梯度方向特征,并对特征的可信度进行加权处理,有效解决了CamShift算法在有颜色相近的干扰目标存在情况下跟踪可能失效的问题。实验表明,该算法提高了跟踪的准确性和稳健性。

关键词: 目标跟踪, MeanShift, 颜色直方图, 梯度方向直方图, 多特征

Abstract: Since the Continuously Adaptive MeanShift (CamShift) tracking algorithm only adopts color as feature, which would result in tracking error, an improved algorithm based on feature fusion was proposed. The presented algorithm detected the moving target automatically using an improved background subtraction method. Object model combined color and gradients orientation features, and weighted the reliability of features. It also overcame the possible CamShift invalidity in the situation of some similar color objects. The experimental results show that the algorithm enhances the reliability and robustness of tracking.

Key words: object tracking, Meanshift, color histogram, gradient orientation histogram, multi-feature