基于二维图像的人脸面部特征点定位与提取  

The Location and Extraction for Features of Human Face Based on Images

作　　者：郭清宇[1] 付志军[1] 

机构地区：[1]中原工学院,郑州450007

出　　处：《中原工学院学报》2011年第5期57-60,共4页Journal of Zhongyuan University of Technology

摘　　要：提出了一种人脸面部特征点定位与提取的方法.利用该方法,在人脸预处理基础上,通过人脸分割、基于水平方向Prewitt算子模版的边缘检测和积分投影技术,可以快速有效地提取眼睛、鼻子和嘴巴等特征点信息.实验表明,该方法可以校正人脸倾斜并具有较高的准确率.An algorithm has been proposed in this paper,which is based on the horizontal Prewitt edge detection algorithm.It includes region growing and integral projection methods,which is proved can extract the face feature point such as eyes and mouth fast and efficiently and it is also proved that face incline can be calibrated and has high nicety rate.

关 键 词：积分投影 人脸特征点 特征提取 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]