基于神经网络和傅氏变换的指纹自动分类系统

AUTOMATIC FINGERPRINT CLASSIFICATION SYSTEM BASEDON PNN AND FT

本文介绍了一个基于两种傅氏变换进行图像增强,用概率神经网络进行指纹分类的指纹自动分类系统,又称AFCS(Automatic Fingerprint Classification System)。该系统运用了一种新的图像增强方法及不均匀的特征表示方法,用PNN(Probabilistic Neural Network)进行分类,实现了指纹的自动分类。

We introduce an automatic fingerprint classification system, where filters are based on two types of Fourier Transform(FT)for enhancing fingerprint images, and a probabilistic neural network classifier is used. We give a new image enhancement method and unequal feature vectors express in this system, and realize automatic fingerprint classification.