Design and Implementation of A Real-Time Face Recognition System

• Main Authors: Yen-Shao Chen, 陳彥劭
• Other Authors: Ming-Yuan Shieh
• Format: Others
• Language: zh-TW
• Published: 2008
• Online Access: http://ndltd.ncl.edu.tw/handle/49968383264391148839

碩士 === 南台科技大學 === 電機工程系 === 96 === This thesis addresses a real-time face recognition system, which can recognize the faces of human beings instantly from the images and also identify the users. The proposed system consists of the face detection and the face recognition subsystems. In the face detection subsystem, it uses the facial color filter to partition the possible facial color areas, and then applies the connected component labeling procedures to localize the objective faces initially. After this, the system will use an ellipse mask to verify where the exact face is in the image, and then normalize these faces blocks for the later faces recognition procedures. In order to recognize the faces, the principal component analysis method is adopted to determine the eigenface space of each face. After constructing the database of the weight vector for all sample faces, one can use the magnitude of the Euclidean distance between the objective and every sample to determine which face sample is the closest and the resultant one. From experimental results, it is seen that the proposed real-time face recognition system has fast and precise recognition abilities. The proposed system in this thesis can be applied to access control systems, security systems, the human-machine interaction systems and service robots.