基于神经网的指纹预分类研究

Fingerprints Preclassification Using Neural Networks

本文提出了一种利用关键点信息而建立的指纹模型，在此基础上设计了适于模型的神经网分类器．最终形成的指纹预分类系统与传统方法相比具有较大的鲁棒性、较高的并行性和较低的预处理要求．

A fingerprint model using key-points information is proposed in this paper, for which a specially tailored neural network is designed to function as its classifier,In this way,a fingerprint recognition system is built, characterized by better robustness, higher parallelism and less preprocessing requirements as compared with the traditional ones.