一种基于连通域分析的卡片指纹图像分割算法  

A Fast and Efficient Algorithm for Slap Fingerprint Segmentation Based on Connected Component Analysis

作　　者：陈小光[1] 王琳[1] 汪周武 

机构地区：[1]北方工业大学图像处理与模式识别研究所,北京100144

出　　处：《北方工业大学学报》2015年第1期1-6,共6页Journal of North China University of Technology

摘　　要：十指指纹卡片捺印是公安部门采集指纹的常用方式.随着自动指纹识别系统(AFIS)的广泛使用,如何快速有效地对卡片指纹进行数字化是一个重要的问题.针对此问题,本文提出了一种基于连通域分析的卡片指纹图像分割方法.对于扫描得到的卡片图像,先进行预处理以获取候选指纹区域,然后通过连通域分析得到有效指纹区域,最后进行指位匹配.在大量指纹卡片上的测试结果表明,该方法能够快速有效地分割出十指指纹图像,非常具有实际应用价值.Ten-print slap cards are widely used for fingerprints by the public security authorities.With the wide application of automatic fingerprint identification system（AFIS）,it becomes very important as how to digitalize the card fingerprints effectively.Aiming to solve this problem,this paper proposes a method based on connected component analysis,which can fast and effectively segment the slap fingerprint images that have been scanned from the ten-print cards.First,apreprocessing procedure is conducted to obtain the candidate fingerprint regions.And then,the valid fingerprint regions are extracted by means of connected component analysis.Finally,each fingerprint image is carefully assigned to its corresponding finger position.The experimental result shows that the proposed method can segment the slap fingerprints effectively and accurately,and is of great potential applications.

关 键 词：自动指纹识别系统(AFIS) 指纹图像分割 指位匹配 连通域分析 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]