摘要
提出了一种基于人眼灰度信息的人脸识别算法。该算法首先根据眼球区域的灰度比它周围区域的灰度低,初步地筛选出可能含有人脸的图像,再把筛选出的图像分割成小图像块,根据图像块的复杂度对人两只眼睛的大概位置进行定位,最后采用居中法,进一步确定人眼的存在,进而确定人脸的确切位置。实验结果表明:此算法具有较高的检测正确率及较快的检测速度。
This article advances a calculation method about face location. Face location is an important basis for automatic human face recognition. In this paper, we have two steps to detect human face. The first step is to find the picture which has the likelihood to have human face by using gray character information. The second step is to use a method to look for the center of eyes. According to experiments, it is concluded that the calculation method has a great efficiency and speed.
出处
《苏州科技学院学报（工程技术版）》
CAS
2005年第2期81-83,共3页
Journal of Suzhou University of Science and Technology （Engineering and Technology）
基金
湖北省自然科学基金项目(No2004ABA068)
关键词
人脸检测
灰度
居中法
automatic human face detection
gray information
center
