| Optical Formula Recognition is the result which the image analysis technology and the traditional character recognition technology, the formula structure analysis technology unify. It becoming more and more important technique. The differences between Optical Formula Recognition and the ordinary OCR are extracting mathematical formula in printed documents, the position of sign recorded as well as the formula structure analysis and the reappearance.It is the most important method that search formula region by the priori knowledge. Through the priori knowledge analysis and the characteristic of sign in the formula, we can find the position of the formula in the article. The emphases of the article are the formula region search, the sign of formula division, the positional information of sign record, the sign of formula recognition and so on. In addition, this article elaborated the application in the search formula region by Chinese character figure information and the priori knowledge of typesetting.Optical Formula Recognition is not only the sign of formula recognition, but also include the extraction of the formula, the record of two dimension structure of the formula and so on. At present, the methods of formula extraction basically are aim at foreign language article. They are not suitable for Chinese article. This article proposed the circulation projection statistics method that finds the corresponding threshold, and then according to the Chinese character figure characteristic discovers the formula region in the Chinese character article. This article comminutes the symbols of the formula and records their position by the outline rectangles of symbols, and then normalized these symbol. Then designed Artificial Neural Network that recognizes these symbol. This article introduced the outline of formula structure analysis and reappearance. This article introduced some commonly used methods. Finally, the algorithm is tested, and confirmed this method will be feasible.
|
PDF Full Text Request