Partial fingerprint recognition using support vector machine

• Main Authors: Vijayaprasad, Perumal, Sulaiman, Md. Nasir, Mustapha, Norwati, O. K. Rahmat, Rahmita Wirza
• Format: Article
• Language: English
• Published: Asian Network for Scientific Information 2010
• Online Access: http://psasir.upm.edu.my/id/eprint/16046/1/Partial%20fingerprint%20recognition%20using%20support%20vector%20machine.pdf; http://psasir.upm.edu.my/id/eprint/16046/; http://scialert.net/abstract/?doi=itj.2010.844.848
• Institution: Universiti Putra Malaysia

In this study, a novel partial fingerprint matching approach that uses global minutiae matching and the support vector machine is presented. Global minutiae-based matching algorithm is used to record the matching pairs and their feature vectors are used to generate a model file which is used for classification. The traditional minutiae-based matching approach is studied as a classification approach by using support vector machine. We used fingerprint verification competition databases for our evaluation. Results from our experiment showed 98.5% of matching is achieved compared to 97.6% obtained using flow network-based approach.