基于邻域搜索的OPTA指纹细化改进算法  

Improved OPTA Algorithm Fingerprint Based on Neighborhood Searching

作　　者：陈刚[1] 陈宁[2] 曾勇[2] 

机构地区：[1]江汉大学数学与计算机科学学院,湖北武汉430056 [2]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《计算机仿真》2012年第9期300-303,共4页Computer Simulation

摘　　要：在指纹细化的研究中,基于OPTA的细化算法及其改进算法普遍存在指纹图像不够光滑,毛刺较多,细化质量不高等现象。为解决上述问题,在OPTA改进算法基础上提出一个新的设计方案。首先重新设计了删除模板和构造模板,然后将该模板设计成一种独特的一维数组查表索引模式,通过计算比较点的邻域值,在表中查找并进行比对判决,即根据邻域的情况来判断一个点是否可以删除。仿真结果表明,用新方法处理后的指纹图像细化完全,骨架接近纹线中心线,光滑无毛刺,且查询速度快,是一种较为成功的细化改进方案。In the study on thinning fingerprint images, there are some problems in the OPTA fingerprint thinning improvement algorithm. For example, the image is not smooth and many burrs are produced in the thinning fingerprint image ; there is not completely thinning. A new OPTA algorithm was proposed in which the delete templates were de-signed and refined as a unique model of table look-up and index by taking OPTA algorithm as the core idea. By com- puting the neighborhood of comparing point, a point was searched and judged in table look-up to check if it is dele-ted. Simulation results show that the processed fingerprint image is smooth and has the complete thinning, the skele-ton is close to the center of grain line without burrs. The proposed algorithm is a successful algorithm of thinning fin-gerprint.

关 键 词：指纹图像细化 邻域搜索 光滑 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]