彩色人脸鉴别特征抽取及识别

Color Facial Discriminate Feature Extraction and Recognition

目前存在的彩色人脸识别方法大多是将彩色图像转换为灰度图像后,采用基于灰度图像的特征抽取与识别算法进行分类识别.本文根据彩色人脸图像的构成特点,对彩色人脸图像的R、G、B三个分量色彩信息进行特征抽取与分析,进而找出适合对彩色人脸图像进行自动识别的鉴别特征.最后,在国际通用的AR标准彩色人脸库中进行了大量实验,验证了本文算法的有效性.

Until recently there have been many color facial recognition methods which almost convert the color image to the gray image, and then conduct the classification and recognition based on the feature extraction and recognition algorithm of the gray image. However, this paper tries to conduct the feature extraction and analysis of the R, G and B three components of the color facial image on the basis of the color facial image feature. Finally, extensive experiments performed on the international and universal AR standard color face database verify the effectiveness of the proposed method.