一种基于频谱分析的指纹纹线距离估计方法  

A method based on spectral analysis for fingerprint ridge distance estimation

作　　者：唐为方[1] 尹义龙[1] 陈旭[1] 

机构地区：[1]山东大学计算机科学与技术学院,山东济南250100

出　　处：《山东大学学报（理学版）》2005年第2期72-78,共7页Journal of Shandong University(Natural Science)

基　　金：山东省自然科学基金重点项目资助 (Z2 0 0 4G0 5 ) ;山东大学青年基金资助项目 (2 0 0 3 -2 0 0 4)

摘　　要：纹线距离是指纹图像纹理结构的本质属性之一,在自动指纹识别中有着重要的应用.已经有人对指纹纹线距离问题进行了研究,但至今尚未形成较为完善的解决方法.采用基于大窗口的频谱分析方法实现了对指纹纹线距离的估计.在FVC2 0 0 0及FVC2 0 0 2部分典型图像上的实验结果验证了本文方法的有效性和鲁棒性.The ridge distance of fingerprints, which is an intrinsic texture property of a fingerprint mage, plays an important role in the automated fingerprint identification. Some researchers have already studied this problem, but so far no satisfying solution has been found. A deep research on the problem was done and the estimation of ridge distance of fingerprint based on spectrum analysis with big window was implemented. Experimental results on some typical fingerprint images from FVC2000 and FVC2002 indicate that this method is effective and robust.

关 键 词：指纹 指纹识别 纹线距离 频谱分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]