Study On Recognition Of Printed Mathematical Expression Based On Characteristic Character

With the development of Internet and computer, people get used to store information through computer. To improve the data transfer speed, we should transfer the text into the electron format that can be processed by computer. To recognize character with computer automatically is the only way to put data into computer rapidly. The highly applied OCR system has a good effect on handwriting recognition and printed text recognition, but there is no good way to recognize mathematical expression, formula recognition becomes a hotspot.Optical Formula Recognition system can be divided into two types: handwriting and printed. We focus on printed formula recognition. A new way is proposed to recognize formula. The whole paper can be divided into four parts:Pretreatment. The pictures we get are usually having noises, it is hard to deal with such pictures directly, so we need do some job at first to make the picture more appropriate to dispose.Segmentation and recognition. Formulas we get usually have many characters, we should make sure which kind of character they are. We propose a method to segment each character from the formula and use another way to recognize it.Structure analysis. We analyze the structure of the formula based on characteristic character that is put forward in this paper. We also give the algorithm in this part. Word EQ syntax is introduced to express the result.Design of software and experiment. We give the structure of the DEMO software and also the main function and the interface. At last we give a demonstration and experiment results.