Character recognition technology of special tickets based on statistical and structure pattern
Zhou Luo-shan · Railway Computer Application · 2010
A characters recognition method of multi-font printed tickers based on statistical and structure pattern was present in this paper. The segmentation technology of special tickets was briefly described, so as to satisfy with requires of post- recognition. It was introduced emphatically the character presort technology based on complex index and four-side code and Chinese characters recognizing method based on feature points. Meanwhile aimed to the problem which existed disturbed point in extracting features, the criteria on basis of stroke directionality was present. Experiments verified, this recognition technology had fast velocity of processing, could identify the staple Chinese characters in tickets, could not be effected by environment noises.