FUZZY LOGIC BASICS

The Markov chain based rank level fusion method has been introduced. The basics of Markov chain have been discussed and its construction mechanism in the context of multimodal biometric rank fusion has been shown. This method demonstrates a number of advantages over other rank fusion approaches in terms of recognition performance. Furthermore, this method satisfies the Condorcet criterion, which is essential in any fair rank information fusion process. In this chapter, another new biometric fusion approach based on fuzzy logic is discussed and hence named as fuzzy fusion for multi biometrics.

Fuzzy fusion method is one of sub-branches of information fusion, which has recently emerged as information consolidation tool. Most fuzzy fusion methods reported in the literature are developed for areas such as automatic target recognition, biomedical image fusion and segmentation, gas turbine power plants fusion, weather forecasting, aerial image retrieval and classification, vehicle detection and classification, and path planning. In the context of biometric authentication, fuzzy logic based fusion approach has recently been used for quality based biometric information consolidation process. The fuzzy fusion method is utilized in multimodal biometric system. The advantage of fuzzy fusion method is that it utilizes both match score and rank information from unimodal biometrics. Also, unlike with traditional systems returning only binary (Yes/No) decision, the level of confidence in recognition outcomes of the multimodal system can be obtained using this method.

Fuzzy logic refers to the theories and technologies that employ fuzzy sets, which are classes with un-sharp boundaries. The idea of fuzzy sets was introduced in 1965 by Professor Lotfi A. Zadeh from the University of California, Berkeley. The core technique of fuzzy logic is based on following four basic concepts:

·         Fuzzy Sets: A fuzzy set is a set with a smooth boundary. Fuzzy set theory generalizes the classical set theory to allow partial membership.

·         Linguistic Variable: A linguistic variable in one which allows its value to be described both qualitatively by a linguistic term and quantitatively by a corresponding membership function (which represents the meaning of the fuzzy set).