| 基于小波变换和梯度矢量的人脸特征点定位 |
| |
| 引用本文： | 陈洁,王典洪. 基于小波变换和梯度矢量的人脸特征点定位[J]. 现代电子技术, 2008, 31(7): 154-157 |
| |
| 作者姓名： | 陈洁 王典洪 |
| |
| 作者单位： | 中国地质大学,机械与电子工程学院,湖北,武汉,430074 |
| |
| 摘 要： | 人脸特征点的定位是自动人脸识别系统和人脸表情识别的重要组成部分,小波变换是近年发展起来的一种有效的信号分析工具。在小波变换基础上,提出了一种基于多尺度梯度矢量的对称变换方法,将此方法应用于人脸特征点定位,不仅减少了计算量,而且提高了定位准确度,同时对光照变化和人脸表情变化也具有很强的鲁棒性。
|
| 关 键 词： | 小波变换 特征点定位 对称变换 融合 |
| 文章编号： | 1004-373X(2008)07-154-03 |
| 修稿时间： | 2007-09-06 |
Location of Facial Feature Based on Wavelet Transform and Gradient Vector |
| |
| CHEN Jie,WANG Dianhong. Location of Facial Feature Based on Wavelet Transform and Gradient Vector[J]. Modern Electronic Technique, 2008, 31(7): 154-157 |
| |
| Authors: | CHEN Jie WANG Dianhong |
| |
| Abstract: | Location of facial feature plays an important part in automatic recognition technology of facial and facial expression,and wavelet transform is a new signal analysis theory developed in recent years.A symmetry transform approach based on multi-scale gradient vector and wavelet transform is proposed.This method applies to location of facial can yield results with higher location accuracy.In addition,the method is also robustness to different illuminant condition and facial expression. |
| |
| Keywords: | wavelet transform feature points location symmetry transform fusion |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
