| 人脸识别系统中的特征提取 |
| |
| 引用本文： | 李华胜,杨桦,袁保宗.人脸识别系统中的特征提取[J].北京交通大学学报(自然科学版),2001,25(2):18-21. |
| |
| 作者姓名： | 李华胜 杨桦 袁保宗 |
| |
| 作者单位： | 北方交通大学计算机与信息技术学院!北京100044 |
| |
| 基金项目： | 国家自然科学基金;69789301; |
| |
| 摘 要： | 研究了人脸识别系统中正面人脸的特征提取 .通过区域增长从人脸图像中分割出人脸 ,再利用边缘检测、Hough变换、模板匹配和方差投影技术可以快速有效地提取出人脸面部器官眼睛、鼻子和嘴巴特征 .实验结果表明本文所采用的方法具有较高的准确率和光照鲁棒性 .
|
| 关 键 词： | 人脸识别 Hough变换 模板匹配 方差投影 |
| 文章编号： | 1000-1506(2001)02-0018-04 |
| 修稿时间： | 2000年12月20日 |
Feature Extraction in Human Face Recognition System |
| |
| LI Hua sheng,YANG Hua,YUAN Bao zong.Feature Extraction in Human Face Recognition System[J].JOURNAL OF BEIJING JIAOTONG UNIVERSITY,2001,25(2):18-21. |
| |
| Authors: | LI Hua sheng YANG Hua YUAN Bao zong |
| |
| Abstract: | This paper deals with frontal facial feature extraction in face recognition system. First human face is segmented from the input image by using region growth. Then by using edge detection, Hough transformation, template matching and variance projection, facial features such as eyes, nose and mouth are located quickly and effectively. Experiment results show that the method we present has a higher accuracy and robust performance to illumination. |
| |
| Keywords: | human face recognition Hough transformation template matching variance projection |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
