期刊名称：International Journal of Innovative Research in Computer and Communication Engineering
印刷版ISSN：2320-9798
电子版ISSN：2320-9801
出版年度：2017
卷号：5
期号：3
页码：4055
DOI：10.15680/IJIRCCE.2017.0503062
出版社：S&S Publications
摘要：Finger vein is a special physiological biometric for identification of individuals based on the physicalcharacteristics and parameters of the vein patterns in the human. The technology is at present in use or development fora wide range of applications, like credit card authentication, security in automobile, employee time and trackingattendance, computer and network authentication, security at end points and automated teller machines. The proposedwork simultaneously acquires information of the finger vein and low resolution finger image and combines these twotechniques using a better score level combination strategy. On analyzing the previously proposed finger-veinidentification approaches in image processing, an approach is proposed that describes its superiority over previouspublished efforts. In this work we have developed and analyzed three new score-level combinations, i.e., Repeated LineTracking , Gabor Filter and Segmentation and SVM to comparatively evaluate analysis with more popular technique ofscore-level fusion approaches to ascertain their effectiveness in the proposed system.
关键词：Finger Vein Recognition; Image Segmentation; Support Vector Machine (SVM)
