Research On Off-line Handwritten Chinese Character Recognition System

Off-line handwritten Chinese character recognition is a challenge in the field of pattern recognition. It will take an important part in many fields of our life, such as letter selecting, check recognition, report form disposing and handwritten manuscript auto-input.The system is mainly used for handwritten manuscript auto-input, The main work on this is as follows:1. In the step of preprocessing, using 3×3 weighted mask for linear smoothing. Especially, different binarizing methods are used according to different paper backgrounds. Such as the examination partition methodology is used on character pictures of blank background and double-threshold methodology is used on the pictures with frame lines. For the sparated character, the system uses linear normalization algorithm to make its size unified.2. Review and summary many kinds of thinning alogrithm for handwritten Chinese characters. Put forword a new thinning algorithm based on edge stripping for handwritten Chinese characters.3. Considering the use of inputting for Chinese characters manuscript, using the systematically segmentation method for the application of this recognition system. Especially, for the single character segmentation, propose a new structural analysis algorithm which is based on storke segment extraction.4. Some statistical and structural feature extraction methods are introduced and the overlapped dynamic grid method is used to extract statistical feature for this recognition system. Meanwhile, put forword the stroke feature extraction algorithm based on thinning method to extract structural feature for the system.5. In the stage of recognition, using an improved double-layer serial classifier structure to save the time about 30%.There are 3000 different Chinese characters as training samples and testing samples, each character has 4 samples. They are divided into two classes according to their styles. One is handwritten printed characters which have many smooth vertical and horizontal strokes. Another is handwritten formal characters which have few running-strokes. The testing results show that the correct rate of the first class is 91% and the second class is 86%.