Interoperability of fingerprint sensors and matching algorithms

Biometric systems are widely deployed in governmental, military and commercial/civilian applications. There are a multitude of sensors and matching algorithms available from different vendors. This creates a competitive market for these products, which is good for the consumers but emphasizes the importance of interoperability. In fingerprint recognition, interoperability is the ability of a system to work with a diverse set of fingerprint devices. Variations induced by fingerprint sensors include image resolution, scanning area, gray levels, etc. Such variations can impact the quality of the extracted features, and cross-device matching performance. This is true even when dealing with fingerprint sensors of the same sensing technology. In this thesis, we perform a large-scale empirical study of the status of interoperability between fingerprint sensors and assess the performance consequence when interoperability is lacking. Additionally we develop a method to increase interoperability in fingerprint-based recognition systems deploying optical fingerprint sensors. A set of features to measure differences in fingerprint acquisition is designed and evaluated. Finally, different fusion schemes based on machine learning are tested end evaluated in order to exploit the designed set of features. Experimental results show that the proposed approach is able to reduce cross-device match error rates by a significant margin