摘要
人脸识别是模式识别领域中一个困难而又具有实际应用价值和广阔应用前景的研究课题．文中用投影法分割人脸图像，采用自适应多主元提取算法提取人脸的主分量特征，设计了一个组合分类器进行人脸的识别．实验结果表明本文所采用的识别算法在特定的部门或场合，对少数人员（目标）身份的准确识别非常有效．
Human face recognition is a very difficult and important problem in the field of pattern recognition,In this paper,the face images are segmented by using pro jection method.Then,An algorithm of adaptive multi-principal components extraction is used to extract the features of face images. Finally,a combinatorial classifier is designed to recognize human faces.Experiment results show that this recognition algorithm is efficient for accurate recognition of the identitiies of a few persons.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
1999年第2期47-51,共5页
Journal of Harbin Engineering University
关键词
人脸识别
主分量特征
神经网络
图像识别系统
face recognition
principal component feature
neural network group
combinatorial classifier
