摘要
提出基于Hessian矩阵特征值比率的增强算法和改进的Gabor滤波特征提取算法相结合的手指静脉特征提取方法。将不同尺度下的Hessian矩阵的特征值应用于静脉增强函数中进行图像增强,然后再通过Gabor滤波进行特征提取,并在Poly U手指静脉数据库对所提算法进行实验验证。结果表明,该算法用于身份识别时的误识率为0. 375%,拒识率为1. 73%,等误率为0. 021%,比目前其他的手指静脉提取算法的识别性能有所提高。
This paper proposed a finger-vein feature extraction method combining an enhancement algorithm based on the ratio of Hessian matrix eigenvalues( ROE) and a feature extraction algorithm based on the improved Gabor filtering. The Hessian matrix eigenvalues at different scales were applied to the vein enhancement function to perform image enhancement and then the feature extraction was conducted through Gabor filters,and the algorithm was validated in the finger vein database of the PolyU. The experimental results show that the algorithm for identification has a false acceptance rate of 0. 375%,false rejection rate of 1. 73% and the equivalent error rate is 0. 021%,which improves the recognition performance of other finger vein extraction algorithms.
引文
[1] KUMAR A,ZHOU Y. Human Identification Using Finger Images[J]. IEEE Transactions on Image Processing,2011,21(4):2228-2244.
[2] WEI P,SHIN J,PARK D. An effective quality improvement approach for low quality finger vein image[C]//International Conference on Electronics and Information Engineering. IEEE,2010:V1-424-V1-427.
[3] YANG J,WANG J. Finger-Vein Image Restoration Considering Skin Layer Structure[C]//International Conference on Hand-Based Biometrics. IEEE,2011:1-5.
[4] PARK Y H,PARK K R. Image quality enhancement using the direction and thickness of vein lines for finger-vein recognition[J]. Int. J. Adv. Robot. Syst,2012(9):1-10.
[5] LEE E C,PARK K R. Image restoration of skin scattering and optical blurring for finger vein recognition[J]. Opt. Lasers Eng,2011,49:816-828.
[6] YU C B,ZHANG D M,LI H B,et al. Finger-Vein Image Enhancement Based on Muti-Threshold Fuzzy Algorithm[C]//International Congress on Image and Signal Processing. IEEE,2009:1-3.
[7] CHO S R,PARK Y H,NAM G P,et al. Enhancement of finger-vein image by vein line tracking and adaptive gabor filtering for finger-vein recognition[J]. Appl. Mech. Mater,2012,145:219-223.
[8] CAO Rongfei,WANG X,WU Zhongke,et al. Cerebrovascular segmentation method based on hessian matrix and clustering[J]. J. Syst. Simul,2014,26(9):2104-2109.
[9] JERMAN T,PERNUS F,LIKAR B,et al. Beyond Frangi:an improved multiscale vesselness filter[C]//Proc. SPIE 9413,Medical Imaging 2015:Image Processing,2015:94132A.
[10]林剑,钟舜聪,张翔.基于Hessian矩阵的手指静脉图像分割[J].机电工程,2016,33(12):1517-1522.
[11] MIURA N,NAGASAKA A,MIYATAKE T. Feature extraction of finger-vein patterns based on repeated line tracking and its application to personal identification[J]. Systems&Computers in Japan,2004,35(7):61-71.
[12] MIURA N,NAGASAKA A,MIYATAKE T. Extraction of finger-vein patterns using maximum curvature points in image profiles[C]//IEICE Trans. Inf. Syst. 2007,E90-D,1185-1194.
[13] VLACHOS M,DERMATAS E. A Region Growing Method Based on Statistical Attributes of Infrared Images for Finger Vein Pattern Extraction[J]. Biomedical Engineering,2012:1-15.
[14] JERMAN T,PERNUS F,LIKAR B,et al. Beyond Frangi:an improved multiscale vesselness filter[C]//Proc. SPIE9413,Medical Imaging 2015:Image Processing,94132A(2015).
[15] PIERPAOLI C,BASSER P J. Toward a quantitative assessment of diffusion anisotropy[J]. Magnetic Resonance in Medicine,1996,36(6):893-906.
[16] FEI Lunke,XU Yong,TANG Wenliang,et al. Double orientation code and nonlinear matching scheme for palmprint recognition[J]. Pattern Recognition,2016,49(C):89-101.