No Detail Found!

Partial Fingerprint Image Enhancement using Region Division Technique and Morphological Transform

Author(s)
A. Ahmad Department of Electrical Engineering, University of Engineering and Technology,Taxila,Pakistan.
I. Arshad Department of Electrical Engineering, University of Engineering and Technology,Taxila,Pakistan.
G. Raja Department of Electrical Engineering, University of Engineering and Technology,Taxila,Pakistan.
Abstract
Fingerprints are the most renowned biometric trait for identification and verification. The quality of fingerprint image plays a vital role in feature extraction and matching. Existing algorithms work well for good quality fingerprint images and fail for partial fingerprint images as they are obtained from excessively dry fingers or affected by disease resulting in broken ridges. We propose an algorithm to enhance partial fingerprint images using morphological operations with region division technique. The proposed method divides low quality image into six regions from top to bottom. Morphological operations choose an appropriate Structuring Element (SE) that joins broken ridges and thus enhance the image for further processing. The proposed method uses SE “line” with suitable angle and radius in each region based on the orientation of the ridges. The algorithm is applied to 14 low quality fingerprint images from FVC-2002 database. Experimental results show that percentage accuracy has been improved using the proposed algorithm. The manual markup has been reduced and accuracy of 76.16% with Equal Error Rate (EER) of 3.16% is achieved.
Publication Details
Page(s) 63-70
DOI DOI not available
Published Journal: Nucleus, Volume: 52, Issue: 2, Year: 2015

Request the full text from PASTIC

Keywords
Partial fingerprint images, Histogram modeling, Morphological transform, Region division
0

Citations

0

Downloads

140

Views