Title :
On the utility of extended fingerprint features: A study on pores
Author :
Zhao, Qijun ; Jain, Anil K.
Author_Institution :
Dept. of Comput. Sci. & Eng., Michigan State Univ., East Lansing, MI, USA
Abstract :
Extended fingerprint features are routinely used by latent examiners in forensic applications. They are now being considered for inclusion in automatic fingerprint identification systems (AFIS), particularly with the adoption of 1000ppi resolution in the Next Generation Identification (NGI) system. Earlier studies on this topic suffered from two limitations: (i) experiments were based on live scan images that are generally of good quality and contain smaller intra-class variations compared to ink images and (ii) the baseline minutiae matcher used to measure the additive value of extended features was not a state-of-the-art matcher. In this paper, we study the utility of pores, one of the most prevalent extended fingerprint features, on rolled ink fingerprint images at both 500ppi and 1000ppi resolution in the NIST SD30 database. The results show that the fingerprint image quality significantly affects the automatic extraction and matching accuracy of pores. Furthermore, the contribution of pores to the overall fingerprint recognition accuracy is miniscule when a COTS matcher is used for 500ppi rolled ink fingerprint images. The fusion between pore matcher and COTS minutiae matcher is a bit more effective on 1000ppi good quality rolled ink fingerprint images. We believe that these results will be useful in the design of next generation AFIS.
Keywords :
feature extraction; fingerprint identification; forensic science; image matching; image scanners; COTS matcher; automatic fingerprint identification system; baseline minutiae matcher; extended fingerprint feature; fingerprint image quality; forensic application; ink fingerprint image; next generation identification system; pores matching; scan image; Bifurcation; Fingerprint recognition; Forensics; Image databases; Image matching; Image resolution; Ink; Law enforcement; NIST; Spatial databases;
Conference_Titel :
Computer Vision and Pattern Recognition Workshops (CVPRW), 2010 IEEE Computer Society Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
978-1-4244-7029-7
DOI :
10.1109/CVPRW.2010.5543239