摘要
为了对核磁共振(MR)检查过程进行科学的管理,文章提出了一种科学的监测系统。该系统主要采用人脸识别技术,首先结合人脸的肤色和特征部位几何分布特征进行人脸定位,然后采用基于奇异值分解和独立成分分析相结合的方法来提取最终特征,可以使分类器的设计更加简洁、有效。实验表明,该文系统识别速度快,并能取得较高的识别率。
The surveillance system advanced in this paper aims at supervising MR cheek-up process.The system mostly adopts face reeoguition technology.The face reeognition technology in this system is firstly proposed using both skincolor feature and feature face component geometry template.Then it extracts the final feature by utilizing the techniques of the singular value decomposition and the independent component analysis.Classifier in this system earl be simplified to make it more compact and effective.The experiment results show that this system has high speed and higher correct recognition rate.
出处
《计算机工程与应用》
CSCD
北大核心
2006年第19期201-204,共4页
Computer Engineering and Applications
关键词
人脸识别
数据库
奇异值分解
独立成分分析
face recognition,database,singular value decomposition,independent component analysis
