基于PCA与MLP感知器的人脸图像辨识技术

摘要

单纯PCA的人脸辨识对未训练样本感知力低的缺点会造成识别率降低。基于PCA与MLP感知器的人脸图像辨识技术是针对该情况提出的，在 PCA 提取有强特征性的人脸图像后，在辨识精度不变的情况下，利用 MLP 感知器对简化的特征属性进行训练，加强监督式规则提取，从而提高识别率。%Primitively PCA face recognition, looking for in the best feature of subspace projection and omitted important caused low recognition rate can be divided into the main considerations. Based on PCA and MLP perceptron face image recognition technology is aimed at the situation, and therefore, after PCA extraction have eigen face, using the MLP perceptron to simplify the property on the characteristics of the training, and then strengthen the supervision and rule extraction, so as to improve the recognition rate.