摘要
符号的大小归一化是数学表达式识别的一种很常见的预处理方法,它对于提高符号的识别率具有重要的意义。本文对表达式符号的大小归一化方法进行了研究。通过比较两种归一化算法的性能,说明在表达式符号的归一化中需要将两种算法结合起来才能满足不同的情况。另外,对于特殊的符号,如长根号需要进行裁剪等归一化预处理,才能获得更好的效果。理论分析和实践证明,本文提出的算法是很有效的,也可作为其他模式识别问题的预处理方法。
The size normalization of notations is a common pretreatment method in mathematical expression recognition, and it has a great significance to enhance the efficiency of notation recognition. In this paper, two methods of size normalization of expression notation are studied. And it turns out that it is necessary to combine these two methods in the process of size nonnalization of expression notations. In addition, for those special notations such as long square root signs, we will not get a satisfying effect unless we normalize it with some methods such as tailoring. It is proved that the algorithm proposed in this paper is very efficient, and it can be used in the preprocessing of the problem of other pattern recognition by theory analysis and experimental practice.
出处
《广西大学梧州分校学报》
2005年第1期79-82,共4页
Journal of Guangxi University Wuzhou Branch
关键词
数学表达式识别
符号识别
预处理
大小归一化
mathematical expression recognition
notation recognition
pretreatment
size normalization