一种改进的彩色人脸鉴别特征抽取方法及自动识别  

A New Color Facial Identification Feature Extraction Method and Automatic Identification

作　　者：高燕[1] 明曙军[1] 刘永俊[1] 

机构地区：[1]常熟理工学院计算机科学与工程学院,江苏常熟215500

出　　处：《常熟理工学院学报》2011年第8期101-105,共5页Journal of Changshu Institute of Technology

摘　　要：针对传统的线性分析方法中都需要的平均样本的共性,提出了基于中间样本的人脸识别.这种方法有效去除了干扰样本对平均样本的影响,并结合彩色人脸识别,提出了基于中间样本的彩色人脸鉴别特征抽取及自动识别方法.最后,在国际通用的AR标准彩色人脸库中进行了大量实验,验证了算法的有效性.Currently face recognition has made some success, algorithms are constantly being improved. According to the common needs of the average sample solution in traditional linear analysis methods, this paper proposes the face recognition based on intermediate samples. This method can remove the influence of average samples to interference samples. Combined with the color of face recognition, the paper proposes color facial identification feature extraction and automatic identification based on the middle samples. Finally, extensive experiments performed on the international and universal AR standard color face database verify the effectiveness of the proposed method.

关 键 词：中间样本 彩色人脸识别 鉴别特征 特征抽取 自动识别 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]