The Research Of Tibetan Character Recognition Technology

Character recognition is set for pattern recognition, image processing and word processing technology with a new technology, is a specific research area of pattern recognition and artificial intelligence. At present, after years of exploration and practice, Western and Chinese printed text recognition is moving toward a practical use. Tibetan character recognition can be widely used in the Tibetan language information processing in various fields, such as the printing press and publishing industry, office automation, data collation and classical studies, digital libraries and so on. In order to better deal with the process of the Tibetan language, facilitate learning and research, information and the books of Tibetan will be saved to computer; therefore, development recognition software is urgently needed. Although there are research of the Tibetan character recognition technology, because Tibetan character has special feature of its structure, the application of traditional recognition methods result the low recognition rate and poor performance. This paper analyzed the character recognition of the concepts and classification, pattern recognition theory and methodology, provided technology proposal of the Tibetan character recognition, using digital image recognition technology highlighted provided a series of rate increase methods in case of interference from the binarization, segmentation, feature extraction, Pattern classification, the correction result and so on. These methods have been successfully used in the printed Tibetan character recognition software and have been proved to be the feasible and efficient.