自动指纹识别系统中关于指纹特征算法的研究  

Research of Minutiae Algorithms on Automated Fingerprint Identification System

作　　者：张志伟[1] 

机构地区：[1]陕西理工学院物理与电信工程学院,陕西汉中723000

出　　处：《吉林化工学院学报》2013年第1期73-76,共4页Journal of Jilin Institute of Chemical Technology

摘　　要：指纹图像的特征提取和匹配在计算机指纹自动识别系统(AFIS)中起着至关重要的作用,直接制约着系统的识别率和识别速度.本文在指纹图像预处理的基础上,对指纹特性算法进行了改进,主要创新点在于简略细化过程同时引入基于二值化图像的特征提取算法.非正向扫描特征提取算法根据指纹的综合特征,沿脊线搜索匹配细节特征点,避免伪特征点的提取.改进后的算法在MATLAB环境下成功应用,实验结果表明,该算法能有效提高识别精度和速度,满足在线验证指纹的实时应用性.The extraction and matching of the fingerprint minutiae plays a vital role in computer Automated Fingerprint Identification System （AFIS）, directly restricts the system recognition rate and speed. Based on the fingerprint image pre-processing, the fingerprint characteristics of algorithm is improved, the main innovation lies in the simple refining process while introducing minutiae extraction algorithm based binarization image. According with fingerprint characteristics, the minutiae extraction algorithm of non-positive scan match minutiae based on ridge tracing, avoid false minutiae extraction. The improved algorithm in MATLAB environment successfully applied, the experimental results show that the algorithm can effectively improve the recognition accuracy and speed, and meet the real-time applications of online fingerprint identification.

关 键 词：指纹识别 预处理 特征提取 特征匹配 算法 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]