期刊文献+

基于Gabor-BLPOC的指关节纹识别算法 被引量:1

Finger-knuckle-print recognition based on Gabor-BLPOC
在线阅读 下载PDF
导出
摘要 指关节纹比手掌特征更明显,针对这种生物特征提出一种基于Gabor-带限相位相关(Gabor-BLPOC)的指关节纹识别算法.首先,使用Gabor滤波器抑制噪声,并采用限制对比度自适应直方图均衡化对指关节纹图像进行增强;其次,使用BLPOC算法提取指关节纹图像的相位特征;然后,通过计算2幅指关节纹图像的互功率谱对指关节纹图像进行校准;最后,再次计算校准后图像的BLPOC,根据2幅图像的互功率谱峰值进行指关节纹图像的匹配.通过在Poly U FKP数据库上的实验表明,所提出算法的等错误率为1.57%,具有更加精确的匹配效果,从而验证了该算法的有效性. Since finger-knuckle-print (FKP)is endowed with more recognizable features than palm character,a novel finger-knuckle-print recognition algorithm based on Gabor and band-limited phase-only correlation (Gabor-BLPOC)is proposed.First,Gabor filter is used to reduce noise,and con-trast-limited adaptive histogram equalization is employed to enhance the FKP image.Secondly,BL-POC is used to extract the phase feature of the FKP image.Then,the cross power spectrum of two FKP images is computed to calibrate the images.Finally,BLPOC of the calibrated images is calcu-lated again,and FKP image matching is achieved according to the peak value of the cross power spectrum.Experiments on the PolyU FKP database are carried out,and it is shown that the proposed algorithm can achieve an equal error rate (EER)of 1 .57%,which verifies the effectiveness of the proposed scheme.
出处 《东南大学学报(自然科学版)》 EI CAS CSCD 北大核心 2014年第6期1121-1125,共5页 Journal of Southeast University:Natural Science Edition
基金 国家自然科学基金资助项目(61201370) 教育部博士点基金资助项目(20120131120030) 中国博士后科学基金资助项目(2013M530321) 中国博士后科学基金特别资助项目(2014T70636) 山东省博士后创新项目专项资金资助项目(201303100) 山东大学自主创新基金资助项目(2012GN043) 高维信息智能感知与系统教育部重点实验室(南京理工大学)基金资助项目(30920140122006)
关键词 指关节纹识别 GABOR滤波器 带限相位相关 互功率谱 finger-knuckle-print (FKP) recognition Gabor filter band-limited phase-only correla-tion (BLPOC) cross power spectrum
作者简介 贲睨烨(1983-),女,博士,讲师,benxianyeye@163.com.
  • 相关文献

参考文献10

  • 1Kumar A, Zhou Y. Human identification using knuckle-codes[ C ]//IEEE 3rd International Conference on Bio- metrics: Theory, Applications, and Systems. Washing- ton, DC, 2009 : 1 - 6.
  • 2Kumar A, Ravikanth C. Personal authentication using finger knuckle surface [ J ] IEEE Transactions on Infor- mation Forensics and Security, 2009, 4( 1 ) : 98 -110.
  • 3Xiong M, Yang W K, Sun C Y. Finger-knuckle-print recognition using LGBP [M]//Advances in Neural Net- works. Berlin: Springer, 2011: 270-277.
  • 4Morales A, Travieso C M, Ferrer M A, et al. Im- proved finger-knuckle-print authentication based on ori- entation enhancement [ J ]. Electronics Letters, 2011, 47(6) : 380 -381.
  • 5Zhang L, Zhang L, Zhang D, et al. Ensemble of local and global information for finger-knuckle-print recogni- tion [J]. Pattern Recognition, 2011, 44(9) : 1990 - 1998.
  • 6Zhang L, Zhang L, Zhang D. Finger-knuckle-print ver- ification based on band-limited phase-only correlation [ C ]//Computer Analysis of Images and Patterns. Mtinster, Germany, 2009 : 141 - 148.
  • 7Li Z C, Wang K Q, Zuo W M. Finger-knuckle-print recognition using local orientation feature based on steer- able filter [ C]//Emerging Intelligent Computing Tech- nology and Applications. Berlin: Springer, 2012 : 224 - 230.
  • 8Mittal N, Hanmandlu M, Vijay R. A finger-knuckle- print authentication system based on DAISY descriptor[ C ]//Proceedings of 12th International Conference on Intelligent Systems Design and Applications. Kochi, Ja- pan, 2012 : 126 - 130.
  • 9Kim M K, Flynn P J. Finger-knuckle-print verification based on vector consistency of corresponding interest points [ C]//2014 IEEE Winter Conference on Applica- tions of Computer Vision. Steamboat Springs, CO, USA, 2014 : 992 - 997.
  • 10Gao G W, Yang J , Qian J J, et al. Integration of multiple orientation and texture information for finger- knuckle-print verification [ J ]. Neurocomputing, 2014, 135 : 180 - 191.

同被引文献6

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部