隐马尔可夫模型在脱机手写数字识别中的应用

The Application of Hidden Markov Models in Offline Handwriting Digital Recognition

摘要: 本文主要用隐Markov模型(HMM)来研究脱机手写数字识别。本文的内容分为三个部分，第一个部分是介绍HMM的基本理论；第二部分是介绍数字图片的预处理和特征提取，这部分属于图像处理方向，并且在手写体数字识别中很重要，特别是提取一个稳定而有效的特征决定着识别是否成功；第三部分是具体地实现这一识别过程，并且用Matlab实现了一个脱机手写数字识别系统。

Abstract: In this paper, we use hidden Markov models (HMM) to study the off-line handwritten numeral recognition. This paper can generally be divided into three parts: the first part introduces the basic theory of HMM; the second part introduces the digital image preprocessing and feature extraction, and this part belongs to image processing, and it is very important in the handwriting recognition, and especially extracting a stable and effective feature determines the successful recognition; the third part implements recognition process, and we use Matlab to obtain an off-line handwriting recognition system.