Hierarchical classification of bank checks using genetic algorithms

This work describes a two-step hierarchical approach that first attempts to assign a query image to a restricted set of classes within the database, and then returns the best matches to each of the selected classes. The proposed method proceeds in two phases: training phase and testing phase. In the training phase, printed patron data are localized and GAs are used to choose the best features, within these locations, that provide an accurate classification. The test phase has a matching strategy that is based on detecting the matching distance between the input image and specific models. Experimental results show that the proposed algorithm is effective and perform well.