Offline Handwritten Chinese Character Recognition Methods

The Chinese Character Recognition System can handle with the Chinese character images, withdraw the essential features of the Chinese characters, and distinguish them after comparing with the template in standard Chinese character libraries. The Chinese character recognition problem which belongs to the classification problem of multi-class pattern, can be divided into the printed Chinese character recognition and the handwritten Chinese character recognition. Further more, the handwritten Chinese character recognition can also be divided into the On-Line handwritten Chinese character recognition and the Off-Line handwritten Chinese character recognition. For the difficulty of recognition, the handwritten Chinese character recognition is harder than the printed Chinese character recognition, and the Off-Line handwritten Chinese character recognition is far harder than the On-Line Chinese character recognition. By now, although the Off-Line Handwritten Number Recognition System has already been applied, the Off-Line Handwritten Chinese Character Recognition System is still in the laboratory. The researches on the Off-Line handwritten Chinese character recognition play an important role in the auto-processing of the Chinese character information and the intelligence of the computer.The thesis is organized as follows.(1) The introduction introduces the background of our study and presents the significance of the Off-Line handwritten Chinese character recognition, and we also point out the current problems and difficult points of the Off-Line handwritten Chinese character recognition in this part.(2) The first chapter introduces the principle and method of the Off-Line handwritten Chinese character mainly includes statistical pattern recognition and structural pattern recognition. Furthermore, a classical method of statistical pattern recognition—project and transform the modulus is introduced.(3) The second chapter illuminates the pre-work of the Off-Line handwritten Chinese character recognition, such as smoothing, binarization, thinning and normalization.(4) The third chapter presents the classification, recognition and the principles of pattern recognition followed by the post-processing technology of the Off-Line handwritten Chinese characters. In this chapter we specify two new methods of the Off-Line handwritten Chinese character recognition. (5) The forth chapter indicates the. tendency of the Off-Line handwritten Chinese character recognition, which is the direction of the author's future work.Basing on the last achievements of the Off-Line handwritten Chinese character recognition we propose two kinds of new methods—the trend of strokes and recovering the order of the Off-Line handwritten Chinese character's stroke, which is helpful to the Off-Line handwritten Chinese character recognition.