Rank level fusion is one of the after matching fusion methods used in multibiometric systems. The problem of rank information aggregation has been raised before in various fields. This chapter extensively discusses the rank level fusion methodology, starting with existing literature from the last decade in different application scenarios. Several approaches of existing biometric rank level fusion methods, such as plurality voting method, highest rank method, Borda count method, logistic regression method, and quality-based rank fusion method, are discussed along with their advantages and disadvantages in the context of the current state-of-the-art in the discipline.
Top2. Review Of Existing Methods
The rank level fusion approach is used in biometric identification systems when the individual matcher’s output is a ranking of the “candidates” in the template database sorted in a decreasing order of match scores (or, an increasing order of distance score in appropriate cases). The system is expected to assign a higher rank to a template that is more similar to the query. Plurality voting method, highest rank method, Borda count method, logistic regression method, Bayesian method and quality based method are reported in the literature to perform rank level fusion in multibiometric system (Abaza & Ross, 2009; Monwar & Gavrilova, 2009; Gavrilova & Monwar, 2008; Monwar & Gavrilova, 2010). All of these biometric rank fusion approaches are discussed in the following subsections of this chapter.
The rank information aggregation problem has been addressed in various fields such as (1) in social choice theory which studies voting algorithms which specify winners of elections or winners of competitions in tournaments, (2) in statistics when studying correlation between rankings, (3) in distributed databases when results from different databases must be combined, (4) in collaborative filtering, and (5) in bioinformatics when gene expression similarity search, meta-analysis of microarray data is needed (Truchon, 1998; Fagin, 1999; Pennock & Horvitz, 2000; Pihur, Datta, & Datta, 2008). The criterion for success is the position of the true class in the consensus ranking, as compared to its position in the rankings before fusion.