【摘要】 视频序列中的目标跟踪效果受到尺度、光照和复杂背景等因素的影响,为此,提出一种结合局部三元模式(LTP)纹理特征和改进型Mean-Shift的目标跟踪方法。首先,为了提高传统Mean-Shift算法对尺度变化的鲁棒性,采用一种尺度和方向自适应的改进型Mean-Shift跟踪器,在均值漂移跟踪框架下估计目标的尺度和方向。另外,通过LTP来提取目标的纹理特征,将其与颜色特征相结合来表示目标模型。实验结果表明:该方法比只采用颜色特征的MeanShift算法更加有效,能处理快速移动物体在目标尺度和方向上的变化。更多还原

【Abstract】 Target tracking in video sequence is affected by scale, illumination and complex background. To solve this problem,a new method of target tracking based on local ternary pattern (LTP) texture features and improved Mean-Shift is proposed. Firstly,in order to improve the robustness of the traditional Mean-Shift algorithm to scale change,an improved Mean-shift tracker with adaptive scale and direction is used to estimate the scale and direction of the target. In addition,LTP is used to extract the texture features of the target and combine them with color features to represent the target model. The experimental results show that this method is more effective than Mean-Shift algorithm which only uses color features,and can successfully handle the changes of fast moving objects in target scale and direction.更多还原