视觉跟踪中模板匹配相似度指标研究

Research on template matching similarity criterion in visual tracking

由于模板匹配中像素点r、g、b颜色值计算存在多对一缺陷,加之背景特征的影响,视觉跟踪中模板匹配往往得不到全局最优解。为此提出模糊隶属度概念和新的相似度指标公式,修正颜色值计算缺陷,相近似颜色值聚类,从而提高视觉跟踪中目标识别能力。这种新指标有效抑制背景成分的影响,同时突出目标特征的权重,改善匹配函数峰值特性,使得搜索目标得到全局最优解,最终实现鲁棒跟踪。实验结果表明,模板匹配具有良好的峰值特性,算法在跟踪目标存在变形、噪声、遮挡时也可以达到比较理想的跟踪效果。

Template matching in visual tracking often can＇ t obtain the global optimal solution because of the defect of pixels r,g,b color＇ s many to one in computation and the influence of background feature. This paper presented a concept of fuzzy membership and a novel similarity measure formula. The proposed method could overcome the defect of color computation and make similar color pixels clustered, consequently it could improve the ability of target recognition. The new criterion could effectively reduce the influence of background feature, and emphasized the importance of target feature. It could improve the peak modality of matching fimction, so the global optimal solution and robust tracking could be easily obtained. Experimental results show that template matching has an excellent peak-like distribution, and tracking algorithm has precision and robustness in the presence of noise, deformation and occlude.