Off-line Unconstrained Handwritten Chinese Character Recognition Of Little Class Set

Optical Character Recognition( OCR) is an important area in pattern recognition research, and handwritten Chinese character recognition is regarded as a famous challenging problem of OCR research. This thesis mainly focuses on off-line unconstrained handwritten Chinese character recognition of little class set.Firstly, a modified nonlinear normalization method based on density equalization of the exact character area is proposed in this thesis, which not only adjusts the relative character stroke position to make the stroke distribution uniform, but also adjusts the stroke thickness to make it consistent so as to narrow the character difference within the same class. Then, a distinguishing method of printed and handwritten legal amount on Chinese bank check was introduced.The Chinese character is made up of strokes or sub-strokes. A fuzzy sub-stroke extraction method is proposed to resolve the unsteadiness because of the unconstrained written fashion. Fuzzy mathematics is the power utility to resolve the conflict between the computer precise calculation and the fuzzy brain thought. This thesis adopts the fuzzy mathematics thought to present the fuzzy sub-stroke statistical feature. The fuzzy sub-stroke statistical feature combines fuzzy division of the character lattice image with the fuzzy sub-stroke property feature. In addition, we present a fast nearest-neighbor search algorithm based on high-dimension features and its application to handwritten Chinese character recognitionFinally, a practical processing system of Chinese bank check later automatic supervising was introduced, including the applying environment primary function, system structure and the main technology used in the sub-system. Adopting the technology presented in this thesis, excellent classifying effect had achieved in handwritten Chinese character recognition aspect of the system.