Improving minutiae extraction in fingerprint images through robust enhancement

Abstract

The paper describes in detail the skeletonbased algorithm for the automatic extraction of fingerprint minutiae. These distinguishing characteristics are used both in the manual and automatic (AFIS-Automatic Fingerprint Identification System) fingerprint matching process for identification purpose. Proposed enhancement algorithm, utilizing multiscale orientation field estimation along with Log-Gabor filters in frequency domain, significantly improves fingerprint image quality, resulting in reduced number of falsely minutiae extracted. Results of experiments conducted on collections of fingerprint images show competitive performance of proposed algorithm.