Font Size: a A A

The Research And Implementation Of On-line Handwritten Mathematical Formula Recognition System

Posted on:2008-03-08Degree:MasterType:Thesis
Country:ChinaCandidate:Y HuFull Text:PDF
GTID:2178360215996911Subject:Curriculum and pedagogy
Abstract/Summary:PDF Full Text Request
With rapid development of computer information technology, people begin to apply them to the field of education. Specifically, handwriting technology and its products are widely used in educational field during the last few years. Recently, the research towards to accurate recognition of hand-written mathematical expressions draws much attention. However, the recognition of hand-written mathematical expressions is generally more difficult when compared to the recognition of hand-written words, since mathematical expressions often characterizes with complex two dimensional linking structures, and many mathematical expressions can be understood in multiple ways. These make the recognition of hand-written mathematical expressions especially difficult.Recognition of on-line handwritten mathematical formula is a complex process. This process can be divided into three steps: symbol recognition, structural analysis, and expression output. Veracity of structural analysis has an important effect on expression output. It is a key step in on-line handwritten mathematical formula recognition.Chapter 1 introduces the background of the research, summarizes the classification of mathematical formula recognition, and raises some technical difficulties. It also introduces current research and products that are related to this topic.In chapter two, technology of pattern recognition technology and the main algorithm of symbol recognition are expatiated carefully.In chapter three, the author focuses on structural analysis algorithm and error correction. At last of this chapter, the author introduces the mathematic editor MathPlayer.In the fourth chapter, the author describes how to design and implement on-line handwritten mathematical formula system. The author advances her opinion about key technologies and realized on PC by programming. Symbol recognition based on statistical method and neural network. It constructs Multi-classifier for saving the time of symbol recognition. The system has learning function in order to extend the recognition ability. In the processing of structural analysis using tree structures to analyze the structure of mathematical expressions. Experimental result indicates the method has better applicability. Application embedded IE component, and the mathematic editor MathPlayer realize formula output. The method is propitious to formula recycle.The end of this paper is the result and the discussion. This part performs a discussion about this article, and takes the result into application. Finally, the future work is proposed based on problems in the system.
Keywords/Search Tags:Recognition of On-line Handwritten Mathematical Formula, Pattern recognition, Symbol recognition, Structural analysis
PDF Full Text Request
Related items