摘要
人脸识别是模式中的一个相当重要却又十分困难的课题。本文利用神经网络（ＮｅｕｒａｌＮｅｔｗｏｒｋ，简称ＮＮ）及主元分析法（ＰｒｉｎｃｉｐｌｅＣｏｍｐｏｎｅｎｔＡｎａｌｙｓｉｓ，简称ＰＣＡ）不同的特性提出了两种人脸识别的模型：ＮＮ＋ＮＮ模型及ＰＣＡ＋ＮＮ模型。理论分析和实验结果表示：这两种新的识别模型可以实现优化特征抽取和自适应识别。
Human face recognition is a very impotant but slso very difficult subject. Through the different feature of Neural Network(NN) and Principle Component Analysis (PCA), this paper proposed two models of face recognition:NN+NN model and NN+PCA model. The oretical analysis and experiment shows that these two models can realize optimized feature extracting and robust recognization.
出处
《小型微型计算机系统》
CSCD
北大核心
1997年第11期36-42,共7页
Journal of Chinese Computer Systems
关键词
人脸识别
神经网络
特征识别
图像识别
Face recognition, Glaobal feature extraction, Neural network(NN), Principle component analysis(PCA)
