Face Recognition Based on The Facial Feature Extraction

• Main Authors: Po-Chou Chen, 陳柏州
• Other Authors: Ji-Chang Tasi
• Format: Others
• Language: zh-TW
• Published: 2013
• Online Access: http://ndltd.ncl.edu.tw/handle/40403078749190547295

碩士 === 義守大學 === 資訊管理學系 === 101 === In today, face detection and recognition have been applied in many domains, such as digital cameras automatically locating the face, determining the number of people watching advertising wall or security systems, etc. Face recognition is an important part in the development of robots. The issue of declining birthrate and increasing aging in our country may be resolved, if we can create a robot that can recognize human faces, and play a role of home care. Common methods of face recognition include principal component analysis (PCA), facial feature extraction, etc. In the previous study, facial feature extraction has obtained accepted result, but the previous extracted features do not yet include the feature of the eyebrow. In fact, the shape of eyebrows has less variation in contrast to other face features, because the mouth and the eyes will be deformed for speaking and blinking, respectively. This is a very good feature of eyebrows, so the feature will be considered in the present study for obtaining the further efficiency of face recognition. The human face database used in the study is created in the natural conditions. That is, some of faces in the face database are suffer from the intensity effects of light. The experimental results show that the method of obtaining eyebrows this study propose can indeed increase the face recognition rate. The rate of correct recognition is nearly 90%.