提出了一种基于DCT提取人脸特征技术和支持向量机分类模型的人脸识别方法。利用离

散余弦变换可提取人脸可识别的大部分信息,而支持向量机作为分类器,在处理小样本、高维数等

方面具有独特的优势,且泛化能力很强,无需先验知识。从ORL 人脸库上的实验结果可以看出,

DCT特征提取是很有效的,且SVM的分类性能优于最近邻分类器,同时提高了整个系统的运算速

度。-A face recognition method based on DCT for face feature extraction and support vector machine classification model. Can extract most of the information face recognition using discrete cosine transform and support vector machine as classifier, has unique advantages in dealing with small sample, high dimension and generalization ability, without prior knowledge. As can be seen from the experimental results on the ORL database DCT feature extraction is very effective, and the SVM classification performance better than the nearest neighbor classifier, while increasing the speed of operation of the entire system.