| With the development of computer technology and integration of information technolgy and curriculum,information-oriented education became more and more concerned.With he use of multimedia teaching,There is a urgent need to replace the traditional keyboard input for handwriting input in order to enhance the efficiency of classroom teaching.However,because own such characteristics of handwritten mathematical formulas as some math symbols are more similar to each other,and some of the more complex mathematical formula exists up/down position, identification of handwritten mathematical formula would be relatively difficult.A handwritten mathematical formula recognition system should be divided into two major on the whole: character recognition and structural analysis.Among them, the character recognition is the basic recognition, the design of character recognition classifier has a direct impact on the recognition rate in recognition system,and structural analysis are the key recognition.In paper, Chapter I introduce research background of the handwritten mathemat ical formula, as well as a number of related commercial products home and abroad, introduced the the general steps on mathematical formula recognition and the work done in this paper.Chapter II focus on the general mathematical symbols preprocessing and feature extraction, preprocessing and distance feature extraction methods proposed in this paper ,use K-L transform to reduce high-dimensional space of feature.Chapter III focus on some common methods of characters recognition, put forward the idea of combination classifier and minimum distance classifier and improved BP neural network algorithm used in this article. On the basis of recognizing mathematical formula,adding the function of studying in order to expand the future ability of recognition.Chapter IV focus on the structural analysis and mathematical formulas Word EQ domain. Mainly introduce how to design and implement the handwritten mathematical formula recognition system, put forward my ideas and programme to complete it.Finally,the paper illustrate the effect of math symbols recognized by neural network BP algorithm and the overall effect of mathematical formula recognition,then put forward the work at the next phase according to the current problems. |
