Research on offline recognition of Chinese hand-written characters

摘要

A method for off-line handwritten Chinese characters recognition is proposed in this work through imitating the recognition process how human understand Chinese characters,.Considering the deformation of handwritten strokes,the basic stroke forms are represented by fuzzy membership degree.Twenty five machine-readable character elements are defined and the fault-tolerance coding is provided for similar character elements.Based on various Chinese character fonts,the judgment rules of the sequence of stroke elements are defined and the stroke order information in characters structure is extracted,and the corresponding code for recognizing the handwritten Chinese characters is generated.In this work,the proposed method is simulated with the HCCORG Chinese characters library.Simulation results verified that the efficient recognition and the adaptability for handwritten deformation are achieved through using the new method for off-line handwritten Chinese characters recognition.