FACE TRACKING BASED ON SKIN COLOR AND α–β–γ FILTER

摘要：

Skin color model is often used for face detection. But it has the problem of over-detection caused by illumination variation and body shadow. Meanwhile, its bigger region of matching search influences the real-time performance of tracking. In order to achieve quick real-time performance of face tracking, this paper presents a new method combined Gauss skin color model with α–β–γ filter, which is to obtain the face initial position through using face motion information and information entropy from frame-to-frame difference image to automatically obtain motion region. Experiments turned out this method can reduce the matching search region of skin color model, increase the running speed and strengthen real-time performance of the system. It also greatly improves the tracking accuracy under disturbing circumstances and the robust performance of tracking algorithm.

关键词： Skin Color Model α–β–γ Filter Face Tracking

作者: Qi Sun Yingchun Liu Yunhua Zhang Donghe Yang

作者单位: Zhejiang Sci-Tech University,Hangzhou,China Zhejiang University of Technology，Hangzhou，China

会议类型: 国际会议

会议名称: China-Ireland International Conference on Information and Communications Technologies 2008(2008 中国-爱尔兰信息与通信技术国际会议 CIICT 2008)

会议地点: 北京

会议语种:英文

页码: 1-4

在线出版日期: 2008-09-26（万方平台首次上网日期，不代表论文的发表时间）