基于角微特征的神经网络识别人脸特征

Facial feature extraction using micro-feature detection based neural network

提出一种利用人脸角微特征几何特性的图像预处理,建立BP神经网络识别人脸特征模型的方法。研究了角微特征提取和具体算法,讨论了BP网络结构的设计,输入、输出层设计和隐藏层节点选取问题。微特征提取,可以降低网络输入维度,对于识别不同角度、不同表情的人脸图像提供了可能性。利用ORL人脸图像数据库做实验,结果表明此方法有效。

A BP neural network approach is applied for facial feature extraction based on a preprocessing technique that utilizes the geometric characteristics of corner mlcro-features. It describes the theoretical background and algorithm of corner detection, discusses the structures of BP network, input, output and hidden layers. The method effectively reduces the dimension of the input to the network and provides possibilities of face recognition in images with various facial expressions and tilting and rotation. By using the ORL database, it shows that the method is effective.