摘要
提出了一种基于单目的复杂环境下强抗干扰性的手势分割算法,使用模糊集合的概念来描述视频流时域和空域上的不同信息,以模糊运算作为信息加工处理的工具。定义了三个模糊集合非背景集、肤色集和模糊手势集,讨论了对模糊集合的腐蚀和膨胀运算。通过对非背景集和肤色集进行模糊运算,得到原始的模糊手势集,然后对原始的模糊手势集进行求精处理。试验结果证明,该文算法实现了对人手的精确分割,且能满足实时性要求。
This paper presents an approach of gesture segmentation with single camera under the complex environment.The time and spatial information are described by fuzzy set,and are processed with fuzzy operation.We define three fuzzy sets：Non-environment Set,Hue Set,and Furry Gesture Set,and then propose fuzzy erosion operation and fuzzy dilation operation.Firstly,we get the crude fuzzy gesture set through fuzzy operation of Non-environment Set and Hue Set.Secondly,the crude fuzzy gesture set is refined.In experiment,the algorithm is applied to hand exact real-time segmentation successfully.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第23期19-23,共5页
Computer Engineering and Applications
基金
国家973重点基础研究发展规划资助项目(编号:G2002CB312103)
关键词
手势分割
模糊集合论
数学形态学
gesture segmentation, Fuzzy Set Theory, Mathematical Morphology
