The Research Of Thinning On Off-line Handwritten Characters Based On Mathematical Morphology

The off-line handwritten Chinese character recognition is an important research subject of pattern recognition.Simultaneously, Chinese characters as the most ancient characters of the world, it possess such qualities as being abundant in quantity, complex in structure, and big amount in similar spellings.therefore, The off-line handwritten Chinese character recognition is also the more difficult topic of pattern recognition. It can be said, the research on off-line handwritten Chinese character recognition not only is of great importance in theoretical value, but also contains significant value in use.Chinese character recognition has generally been divided into:pre-operation, feature extraction, classification methods and reprocessing. While,the greatest impact on the recognition performance is the pre-operation, feature extraction and classification methods. In this paper, the focus is the pre-processing stage of research on the off-line handwritten Chinese character recognition.The major steps of pre-processing on off-line handwritten Chinese character recognition can be divided into:Binarization,Smoothing, Image segmentation, Normalized,Thinning and so on. In this paper, these steps were analyzed, proposed a improved method of thinning algorithm based on Mathematical Morphology, by reviewing and summing up a number of commonly used thinning algorithm.This improved method not only has the superiority of high recognition rate in the field of handwritten Chinese character recognition,but also improves the speed thinning.It achieves a good recognition effect.This topic deep researched Mathematics Morphology thinning method,gave full consideration to small character set handwritten Chinese character recognition, suitabled and target-oriented changes original structural element, proposed a algorithm of simpling array based on changing the scanning sequence—The thinning algorithm on fast structural element match. In this algorithm,the original array which containes 256 elements be reduced to the array which containes only 54 elements.so, every time,it will save a lot of scanning time when scanning array,and improve the efficiency of thinning.at the same time, the correct recognition rate and recognition speed have been greatly increased through a large number of test to experimental sample collected and effective adjustment structural element.The application of this algorithm has very strong pointedness.In some specific applications, it will play powerful effect,and it is a reparation to the defect of Chinese character recognition rate is not very high when the large character set handwritten Chinese character recognition system used in specific field. It provides some reference value for the research of future off-line handwritten Chinese character recognition.