Call of Papers for Current Volume ********************OnLine Paper Submission for Current Volume

Volume & Issue no: Volume 3, Issue 5, September - October 2014

____________________________________________________________________________________________________

Title:
A New Approach to Biometric Person Identification from Retinal Vascular Patterns
Author Name:
Ritesh Kumar Gharami and Malaya Kumar Nath
Abstract:
Abstract In this paper, a novel method for biometric person identification has been proposed by using retinal vascular patterns present in the retina. Retinal vascular patterns crossing the optic disk in the eye is a unique feature and varies from person to person. Here, singular value decomposition (SVD) is used for feature extraction and a 7-State hidden Markov model (HMM) is used for identification. The blood vessels patterns are segmented from the green-channel fundus image using Gaussian matched filter based segmentation method as Gaussian matched filter segmentation method segment the blood vessels more accurately than the other methods.The segmented fundus image is then manually cropped at the optic disc region. A top-down observation sequence of overlapping sub-image blocks is considered for finding SVD coefficients of each retinal image.A small number of SVD coefficients are used as features associating each block, which makes the training andrecognition process faster. The proposed system has been evaluated on the Digital Retinal Images forVessel Extraction (DRIVE) database, Glaucoma database and Automated Retinal Image Analysis (ARIA) database. The system achieved a recognition rate of 95% when examined on 40 fundus images from DRIVEdatabase, a recognition rate of 83.33% for 18 glaucoma affected fundus image from Glaucoma database, and a recognition rate of 100% for the 132 fundus images from ARIA database. The proposed methodper forms fast and reliable identification even for retina affected by pathologies or glaucoma. This methodalso works well for the fundus images affected by salt and pepper noise. Keywords:- retinal vascular pattern, fundus image, SVD, HMM, person identification
Cite this article:
Ritesh Kumar Gharami and Malaya Kumar Nath , " A New Approach to Biometric Person Identification from Retinal Vascular Patterns" , International Journal of Emerging Trends & Technology in Computer Science (IJETTCS) , Volume 3, Issue 5, September - October 2014 , pp. 191-195 , ISSN 2278-6856.
Full Text [PDF]                           Back to Current Issue

NOTE: Authors note that paper cannot be withdrawn at any condition once it is accepted. The Team of IJETTCS advise you, do not submit same article to the multiple journals simultaneously. This may create a problem for you. Please wait for review report which will take maximum 01 to 02 week. 

 

Contact us


International Journal of Emerging Trends & Technology in Computer Science (IJETTCS)
ISSN 2278-6856
Frequency : 6 Issues/Year


E-mail: editor@ijettcs.org