MATLAB数字识别系统的开发研究

摘要：近年来，科技迅速发展，尤其是在计算机技能和图像处理技巧方面，人们也顺理成章的对模式识别方面的技术提出了更高的要求，针对数字识别技术的应用也更加广泛。比如，如今办公用品中几乎是必备的一种器材—扫描仪，它的工作原理就是把手写的文字草稿扫描成图像，在这个过程中，就是完成了识别手写体使之成为计算机能够辨别的字符。本文仔细介绍了手写体数字识别的方式和进程，方式主要囊括了手写体数字识别的训练和过程。首先我们会对准备输入图像进行一些预处理，使之能够提出有效的像素值，方便计算机的采取，然后再设计一种神经网络并进行训练，之后对训练产生的结果和预先期望的结构进行反复比较，并根据比对的差异性对神经网络的权值进行部分微调，至此训练好的神经网络才能够被我们所得到。下一步才是选择测试样本，进行仿真测试。本文一方面通过MATLAB软件实现手写体的数字识别技术，另一方面，并制作GUI界面上，实现人机互动，使设计更加的灵活直观，该文开发的手写体识别系统可以在画板上手写数字并保存，然后对该手写体数字进行识别。70980

毕业论文关键词：图像处理；数字识别；MATLAB；GUI界面；

Research and Development of Digital Recognition System

Abstract: With the development of computer technology and image processing technology in recent years, people have more requirements for pattern recognition technology. The application of digital recognition technology is becoming more and more extensive. For example, the scanner scans the printed body or handwritten document into images, And then identify the corresponding computer can be directly processed characters. This paper focuses on the method and process of handwritten numeral recognition. The method is mainly composed of handwriting digital recognition training process and recognition process. Firstly, the preprocessing of the input image is put forward, the effective pixel value is put forward, a neural network is designed and trained. The results of the training and the expected structure are compared, and some weights of the neural network are modified according to the contrast result. Train the neural network, and select the test sample, the simulation test. This paper on the one hand through the MATLAB software to achieve handwriting digital recognition technology, on the other hand, and the production GUI interface, the realization of human-computer interaction, so that the design more flexible and intuitive, the development of the handwriting recognition system can handwritten figures on the drawing board and save , And then the handwritten numeral is identified.

Keywords: Image processing; digital recognition; MATLAB; GUI interface;

摘要 i

Abstract ii

目录 iii

1 绪论 1

1.1 课题研究背景和意义 1

1.2 模式识别技术的发展及现状 1

1.3 OCR技术的发展及现状 2

1.4 论文结构和内容 2

2 数字识别的基本方法 4

2.1 图像预处理技术 4

2.1.1 MATLAB数字识别系统的开发研究:http://www.youerw.com/zidonghua/lunwen_80493.html