A comparative study of Biometric techniques for person identification

Gupta, Divya and Prabhu , Srikanth and Hegde, Govardhan and Kini, Gopalakrishna N (2013) A comparative study of Biometric techniques for person identification. In: 2nd International conference on Computer Engineering and Mathematical Sciences, 5 and 6th Dec 2013, Malaysia.

[img] PDF
ICCEMS 2013 Malaysia.pdf - Published Version
Restricted to Registered users only

Download (508kB) | Request a copy

Abstract

The process of separating foreground from background is called segmentation. In this way the features for fingerprint biometric can be extracted. In this paper emphasis has been laid to correlate fingerprints with human retina, using the feature, bifurcations and minutiae points. Segmentation on images is performed by calculating ridges or morphology. Ridges are those areas in the images where there is sharp contrast in features. The features are targeted using morphology, using structuring elements. Structuring elements are of different shapes like disk, line which is used for extracting features of those shapes. When segmentation was performed on retinal images and finger print images, problems were encountered during image pre processing stage. Edge detection techniques have been deployed to find out the contours of the retinal images and finger print images. After the segmentation has been performed, it has been seen that artifacts of both the biometric images have been minimal when ridge based segmentation technique was deployed. Also parallel processing methods have been applied, to extract the bifurcation points and minutiae points, which has given 95 % accuracy for the images.

Item Type: Conference or Workshop Item (Paper)
Uncontrolled Keywords: Finger Print, Retina, Minutiae, Laplace, Gaussian.
Subjects: Engineering > MIT Manipal > Computer Science and Engineering
Engineering > MIT Manipal
Depositing User: MIT Library
Date Deposited: 27 Dec 2013 05:42
Last Modified: 27 Dec 2013 05:42
URI: http://eprints.manipal.edu/id/eprint/138110

Actions (login required)

View Item View Item