基于三方向图的多尺度平滑指纹奇异点检测

详细信息    查看全文 | 推荐本文 |

英文篇名：Multi-scale smooth and fingerprint singularity detection based on three-directional image 作者：李海燕 ; 程龙 ; 宗容 ; 陈建华 英文作者：LI Haiyan;CHENG Long;ZONG Rong;CHEN Jianhua;School of Information Science and Engineering,Yunnan University; 关键词：Gabor滤波器 ; 频率估计 ; 图像增强 ; 方向图 ; 像素交汇块 ; 多尺度平滑 ; 奇异点 英文关键词：Gabor filter;;frequency estimate;;image enhancement;;directional image;;intersection blocks of pixels;;multi-scale smoothing;;singularity 中文刊名：HZLG 英文刊名：Journal of Huazhong University of Science and Technology(Natural Science Edition) 机构：云南大学信息学院; 出版日期：2019-03-13 16:21 出版单位：华中科技大学学报(自然科学版) 年：2019 期：v.47;No.435 基金：国家自然科学基金资助项目(61561050,61661050);; 云南省高校重点实验室建设计划资助项目 语种：中文; 页：HZLG201903017 页数：6 CN：03 ISSN：42-1658/N 分类号：103-108

摘要

为解决低质量指纹图像定位奇异点位置偏移及定位区域过大的问题,提出了一种基于三方向图的多尺度平滑奇异点检测算法.首先,计算指纹图像的方向场和频率场估计并用Gabor滤波器对指纹图像进行增强.然后,计算指纹图像中脊线上像素点的方向并对其进行三方向划分得到指纹脊线方向图.最终,去掉噪点并填补非脊线区域,同时检测多个平滑尺度下的奇异点位置并通过信息融合剔除伪特征点,精确定位奇异点.对FVC2002,FVC2004和FVC2006指纹图像库进行实验,结果表明:该算法对低质量指纹图像奇异点检测的准确率有明显提升,奇异点检测的总准确率达92.35%,可将奇异点精确定位在2×2像素区域内.
        In order to solve the issues of singularity position offset or inaccurate singularity detection in low quality fingerprint images, a method on multi-scale smooth and fingerprint singularity detection based on the three-directional image was proposed.Firstly,the directional field and the frequency estimation were calculated and a Gabor filter was applied to enhance the low quality fingerprint image.Subsequently,the pixel direction of fingerprint ridge was computed,then the directional image of fingerprint ridge was obtained by dividing the directions into three parts.Finally,after eliminating the noise and completing the non-ridge area,singularities under multi-scales of smoothing were detected.Furthermore,information fusion was applied to eliminate the false feature points to accurately locate the singularities.Experimental results on FVC2002,FVC2004 and FVC2006 show that the accuracy of singularity detection obtained by the proposed method is improved and the accuracy ratio is 92.35%.Furthermore,the proposed method can precisely locate the singularities in a 2×2 pixels area.

引文

[1]KARU K,JAIN A K.Fingerprint classification[J].Pattern Recognition,1996,29(3):389-404.
    [2]JIN C,KIM H.Pixel-level singular point detection from multi-scale Gaussian filtered orientation field[J].Pattern Recognition,2010,43(11):3879-3890.
    [3]PARK C H,LEE J J,SMITH M J T,et al.Singular point detection by shape analysis of directional fields in fingerprints[J].Pattern Recognition,2006,39(5):839-855.
    [4]YANG J,LIU L,JIANG T,et al.A modified Gabor filter design method for fingerprint image enhancement[J].Pattern Recognition Letters,2003,24(12):1805-1817.
    [5]TURRONI F,MALTONI D,CAPPELLI R,et al.Improving fingerprint orientation extraction[J].IEEE Transactions on Information Forensics and Security,2011,6(3):1002-1013.
    [6]BIAN W,DING S,XUE Y.An improved fingerprint orientation field extraction method based on quality grading scheme[J].International Journal of Machine Learning and Cybernetics,2018,9:1249-1260.
    [7]HUANG C Y,LIU L,HUNG D C D.Fingerprint analysis and singular point detection[J].Pattern Recognition Letters,2007,28(15):1937-1945.
    [8]CHUA S C,WONG E K,TAN A W C.Fingerprint singular point detection via quantization and fingerprint classification[J].World of Computer Science and Information Technology Journal,2015,5(12):172-179.
    [9]BAZEN A M,GEREZ S H.Systematic methods for the computation of the directional fields and singular points of fingerprints[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2002,24(7):905-919.
    [10]GOTTSCHLICH C.Curved-region-based ridge frequency estimation and curved Gabor filters for fingerprint image enhancement[J].IEEE Transactions on Image Processing,2012,21(4):2220-2227.
    [11]SILVA A,SILVA P,BATISTA L,et al.A novel approach for fingerprint singularities detection[C]//Proc of 2017Workshop of Computer Vision(WVC).Natal:Brazil,2017:102-107.