Research on decision information system of off-line handwritten Chinese character recognition based on variable granular theorem

Based on the rough set theory and variable granular theorem, a kind of decision information system of off-line handwritten Chinese character recognition is constructed. The performance index entropy in this decision information system is defined in this paper. According to the definition of attribute relative significance, the feature attributes are classified, that the representation of feature attributes in single layer and single granularity become multiple layers and multiple granularities, and a hierarchical reduction approach is designed. The weak points of the single granularity that fixed and reduction is not fine are made up. Based on the D-S evidence theory, the decision rules after reduction are fused, so that the recognition rate can be improved. The experiment results show that the method is feasible and effective.