摘要
提出了基于HMM的人脸识别方法.根据人脸的特点建立HMM模型,然后采用奇异值分解抽取观察序列.通过对观察向量的仿真结果表明,这种方法在提高识别率和缩短计算时间方面都有很大进步.
The paper proposed a method of person face recognition based on the HMM. Establishes the HMM according to the person face＇s characteristic, then extraction observation sequence with the singular value decomposition. Through the simulation result of observe vector, this method is very greatly progressive for the reduction computing time and enhancement recognition rate.
出处
《河南科学》
2008年第10期1254-1256,共3页
Henan Science
关键词
HMM
识别
奇异值分解
观察向量
HMM
recognition
singular value decomposition
observe vector
