Auriculocephalic Sulcus Localization for Fast Semi-3D Face Vertical Pose Recovery and Its Android Embedded System Implementation

• Main Authors: Chun-Xuan Liao, 廖椿玄
• Other Authors: Chian C. Ho
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/05915837930736624483

碩士 === 雲林科技大學 === 電機工程系碩士班 === 98 === Face recognition technology is popular in biometric field because it can provide noninvasive and convenient features. However, a face recognition system seldom gets frontal-face images so that many facial pose recovery methods has been developed. Fast semi-3D face vertical pose recovery is the only way to solve the declination problem in real time without wasting the space of database and knowing the declination angle. Fast semi-3D face vertical pose recovery method depends on both accurate localization of eyes and auriculocephalic sulcus. However, we can get the eyes locations easily through Haar-like features and Adaboost algorithm, but it is not suitable to detect auriculocephalic sulcus in the same way because there are few discriminative features around the auriculocephalic sulcus. Therefore, this thesis proposes contoured skin detection method for auriculocephalic sulcus localization. On the other hand, this thesis also implements the auriculocephalic sulcus localization beased on contoured skin detection method onto Google Android embedded platform. Experimental results show the proposed contoured skin detection method is more useful to localize the auriculocephalic sulcus than conventional skin detection method or edge detection method.