FUZZY FUSION OF BIOMETRIC

Fuzzy logic is indeed one of the fascinating areas on the edge between cognitive science and decision making. Utilizing principles of fuzzy logic for information fusion allows emulating the abstract reasoning and complex human intelligence processes by the means of a range of values between yes or no, or true and false. In a machine language, it is a difference between 0 or 1, and in biometric terms—between accepted and refused identity, or granted or denied access to a secure premises or a facility.

One of the first works in this domain which has significance for decision-making is 1999 work by Solaiman et al. The authors proposed a fuzzy-based multi sensor data fusion classifier for to be used in a geo-spatial and remote sensing domain for land cover classification. Their classifier provided a tool for integration of multi sensor and contextual information. The authors introduced the Fuzzy Membership Maps (FMMs) to represent different thematic classes based on a priori information obtained from sensors. The FMMs were next iteratively updated using spatial contextual information. The fuzzy logic allowed their proposed classifier to integrate multi sensor and a priori information.

In the experiment, the same two datasets which were used in the experiments involving Markov chain-based rank fusion are used. Here, comparison has been made on fuzzy fusion approach with unimodal matchers, with the rank fusion approaches and with Match score and decision fusion approaches.