1. UKM Learning and Research Repository
2. Master Thesis / Tesis Master
3. Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Title:	Finger vein visualization using an unbiased detector of curvilinear structures
Authors:	Zeinab Fardkakaie ( P62455)
Supervisor:	Riza Sulaiman, Prof. Dr.
Keywords:	Information security Personal Identification Number (PIN) Universiti Kebangsaan Malaysia -- Dissertations
Issue Date:	21-Oct-2013
Description:	In the current information age, people pay more attention to the science and technology's information security developments to overcome some of the difficulties of the traditional detection methods such as username and password verifications. Personal Identification Number (PIN), smart cards and passwords are slowly being replaced by biometric verification methods. Biometrics is the science and technology of identifying and/or authenticating individuals using their physiological and/or behavioral specifications. Physiological biometrics consists of the face, eye (retina or iris), finger (fingertip, thumb, finger length or pattern), palm (print or topography), and the geometry of the back of the hand’s vein pattern or thermal images. Behavioral biometrics consists of voiceprints, handwritten signatures, and keystroke/signature dynamics. In comparison to the other traditional biometric methods Face Recognition System is sensitive to Illumination, Pose, and Facial elements and can also be easily spoofed by a static photo or moving video. In the last decade, biometric technology has become an important and high requirement for security access systems. It also seems a logical and more secure way of verifying individuals. Hence, the overall aim of this research is to propose finger vein authentication for smart access application. In particular The research focus our attention on; i) B- Huang method was used to normalize images to reduce the irregular distortions caused by variance of finger pose as well as this work could detect and remove noise from captured image using anisotropic nonlinear diffusion that is often essential step in image pre-processing; ii) to extract feature, the vessel enhancement conceive as a filtering process that looking for geometrical structures which can be regarded as tubular, First for each pixel the second derivative response kernel is collected and formed into a Hessian matrix which undergoes eigen analysis. From there, the best eigen value response along with the corresponding eigenvector is chosen on different scales. Using the resulting principal directions of linear segment pixels enable us to identify and enhance vessels the proposed function by Frangi to distinguish vessel from background and differentiating line-like from blob -like ; and iii) in final step, neural network was applied to match. Experimental results show that the proposed method is capable of extracting finger veins images which are reliable and effective. Further experiments illustrated that the accuracy rate was 99.9% which means the method is very effective as a means for personal authentication. To prove this claim this approach compared with two methods maximum curvature and line tracking in same condition of preprocessing and matching steps also at same data bases for them, the results were shown 99.7% and 94.6% for maximum curvature and line tracking respectively,Certification of Master's/Doctoral Thesis" is not available
Pages:	99
Call Number:	TA1634.F346 2013 3 tesis
Publisher:	UKM, Bangi
Appears in Collections:	Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.