复合指纹核心点定位方法  

A Compound Algorithm for Locating Fingerprint Core Points

作　　者：张柯[1] 孙刘杰[1] 

机构地区：[1]上海理工大学光电学院,上海200093

出　　处：《计算机仿真》2009年第1期240-242,338,共4页Computer Simulation

基　　金：上海市重点学科基金印刷出版(P0501)

摘　　要：对于指纹核心点定位的诸多算法,通常都存在漏检、偏检以及耗时过长等缺点。提出了一种复合指纹核心点定位算法,首先运用方向图信息进行粗定位,确定核心点的候选区域,然后通过Poincare索引法精确定位核心点。对于粗定位改进了传统算法的检测条件,定位出的核心块既精确而且面积较小,大大减少了下一步精确定位核心点的运算量,且对Poincare索引法检测中心点存在的不足,引入方向一致性度量加以弥补。经大量实验证明算法适应性较强,很好的解决了漏检等问题。Many algorithms for locating fingerprints core points, often have some disavantages, such as leaking, deviation and time consuming. This article puts forwad a compound algorithm for locating fingerprints core points. First, getting the candidate regons by directional image, and then locating the core points accurately by Poincare Index. And this article improves the conditions of detecting core points, the regions detecting by this way are very accu- rate and small, this will decrease the operations in next step. And introduing directional consistency in order to make up for the deficiency of Poincare Index. Lastly, many experiments prove that this algorithm has strong adaptability, and can solve many problems.

关 键 词：指纹图像 核心定位 方向图 改进 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]