人脸特征提取与识别技术研究  

Technology of Human Face Feature Extraction and Recognition

作　　者：魏冬冬[1] 谌海新[1] 聂铁铸[2] 

机构地区：[1]国防科技大学电子科学与工程学院信息工程系,湖南长沙410073 [2]酒泉卫星发射中心,甘肃酒泉732750

出　　处：《计算机与现代化》2007年第3期69-72,76,共5页Computer and Modernization

摘　　要：人脸识别是模式识别和人工智能领域的一个研究热点,近年来,特征提取与识别作为人脸识别系统的关键环节得到了更为广泛和深入的研究。本文首先简要回顾了人脸识别技术的研究背景及发展历程,对目前人脸特征提取与识别的主要方法进行了分类总结。通过对各种特征提取与识别方法的分析与比较,总结了当前存在的技术困难,并展望了今后的研究趋势。The human face recognition is a hot research area in the field of pattern recognition and artificial intelligence. In recent years, as a key step of the human face recognition system, feature extraction and recognition have been more widely and thoroughly researched. Firstly, research background of face recognition and its development are reviewed briefly, the existing main methods of the human face feature extraction and recognition are classified. In addition, through analysis and comparison of various methods, the paper summarizes the current technology difficulties and forecasts the future research tendency.

关 键 词：人脸识别 特征提取 弹性图匹配 人工神经网络 支持向量机 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]