指纹认证的统计学方法及滤波器的参数分析

A Statistical Approach to Fingerprint Verification and Analysis of Filter Parameters

本文讨论了一种基于灰度值进行指纹认证的统计学方法 ,详细阐述了搜索中心点、确定特征提取区、调整特征区灰度、使用方向滤波器提取指纹特征等关键算法的细节 ,并对算法进行了理论分析 .此外 ,我们重点对滤滤算法中的滤波器模板尺寸、滤波方向娄等一些可调参数进行了优化研究 ,并给出了试验结果 .

In this article,a statistical way of verifying fingerprints based on the gray-scale is proposed.We detailedly expound each step in the process of fingerprint verification which is composed of core searching, features area locating,gray scale normalizing and features extracting.Theoretical analysis of the proposed algorithm is also made in this article.Furthermore,we attach more attention to the optimization of some tunable parameters such as the size of the filters and the number of directions of filters.This optimization of the proposed algorithm turns out to be successful through analyzing a lot of experimental data.