欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
为了提高在目标被遮挡及目标颜色发生较大变化情况下CamShift算法跟踪的稳健性,文中在目标建模阶段不仅利用了目标的颜色信息,而且还加入了目标的空间信息,建立了“空间直方图”,从而提高算法的精确性及抗干扰能力;同时,依据径向距离理论提出了一种新的模板更新策略,进一步增强了算法的鲁棒性.实验结果证明,改进方法是有效的.
英文摘要：
Classic CamShift tracking algorithm limits the performance when the object is occluded or changes its color in large range. In order to improve the algorithm＇s stability, we introduce a new concept of a spatiogram, which includes not only the color information but also spatial information about the object. And using of the theory of radial-distance, we raise a new method to update the target model continually. Experimental results, tested on various sequence with scenes partial occlusion and rotations in depth, indicate that the two improvements enhance the precision and stability of the original algorithm.
同期刊论文项目
同项目期刊论文
期刊信息
位置：