人脸检测系统研究及实现

Research on Human Face Detection System

人脸检测识别因其广阔的应用前景而成为当前模式识别与人工智能领域的研究热点之一。文章对带有人脸信息的24住真彩色BMP文件,通过基于肤色信息的彩色图像分割方法,利用预定的肤色建模知识进行人脸区域检测及区域标定。在此基础上,利用给定的眼、嘴先验信息进行眼、嘴的匹配捡测,并最终实现双眼中心以及嘴的标定,从而实现人脸及其特征信息的检测。

Face recognition and detection has become a hotpot in artificial intelligence research for its importance in surveillance and security, telecommunications, iatrology, digital libraries, video meeting, and human-computer intelligent in-teractions.This paper introduces a system which can deal with a 24 bit multicolor image.With the division of the image based on complexion information,and the help of pre-established information of complexion model,the system can detect the face area and locate it on the image.After that,the system matches the eyes and mouth according to the given information.Finally, the system locates the center of eyes and mouth within the image.