Face and facial components detection in complex background color images

作     者：Kim, Jin-Tae Kim, Dong-Wook Kim, Jiekwan Lim, Hankyu Song, Ho-Keun 

作者机构：Department of Computer and Information Science Hanseo University Seosan-si Chungcheongnam-do Korea Republic of Department of Information and Communication Engineering Jeonju University Jeonju City Korea Republic of Department of Industrial and Systems Engineering Changwon National University Kyongnam Korea Republic of Department of Multimedia Engineering Andong National University Andong City Korea Republic of 

出 版 物：《WSEAS Transactions on Circuits and Systems》 (WSEAS Trans. Circuits Syst.)

年 卷 期：2005年第4卷第12期

页      面：1805-1810页

核心收录：

主　　题：Color image processing 

摘      要：The extraction of face and facial components in images is a preprocessing for face recognition. In this paper, a new method for the extraction of face and facial components in color images is proposed. For efficient extraction using image processing, we analyze and improve detecting facial components that are the basis of an image characteristic. The proposed algorithm includes face region detection without the effect of skin color hairs, eye region detection with weighted template matching, eyebrow region detection using a modified histogram, and mouth region detection using skin characteristics. This method provides better accuracy in the extraction of face region and facial components compared with the existing methods and several faces in the same image can be extracted.