Fingerprint classification technique using gradient-based method for singular points detection

Abstract EN

The recent advances in technology caused increased of fingerprint recognition systems uses.

To identify or verify the people, usually the fingerprint recognition systems composed of a huge database that contains a large number of Fingerprints.

Fingerprint classification is the efficient technique used to improve the performance of fingerprint recognition systems by reducing mathematical complexity as it reduces the response time of the system.

This paper proposed a fingerprint classification technique.

A rule based classifier used to classify fingerprints based on singular points detection.

The orientation field estimation using Prewitt operator in gradient computing was applied with the classifier.

In this study, the conducted experimental used the Fingerprint Verification Competition (FVC 2004) database (Set B) with 320 fingerprint images.

The results showed that the maximum accuracy achieved was 53.75% and the minimum accuracy achieved was 16.25%.