基于径向基函数网络人脸识别的研究

The Research on Face Recognition Based on the Radial Basis Function Network

用人脸信息来识别和辨认一个人类个体,是计算机视觉和模式识别领域中的一个研究热点。本文提出了一种基于径向基函数网络(RBFN)识别人脸的方法,使用主分量分析(PCA)技术降低样本维数,并用生成图像(SI)技术改变人脸的姿态,以增加学习样本数。用标准人脸库ORL进行实验,表明人脸识别效果有大幅度的提高。

Using face feature information to recognize or verify an individual is a research hotspot in computer vision and pattern recognition fields. In this paper we describe a face recognition method based on Radial Basis Function Network (RBFN) architecture. The proposed method reduces the dimension using Principal Component Analysis (PCA) approach. In order to increase the number of samples for training, we augment the training set with additional, synthetically-generated face image. The experiments carried out on ORL face database set. Results of these experiments show the improvement of recognition performance.