帮助 本站公告
您现在所在的位置:网站首页 > 知识中心 > 文献详情
文献详细Journal detailed

基于图像识别技术的动态几何系统的开发研究
Research on Dynamic Geometry System Based on Image Recognition Technology

导  师: 李传中

学科专业: 040102

授予学位: 硕士

作  者: ;

机构地区: 广州大学

摘  要: 随着计算机和信息技术的发展,网络学习资源库的发展如雨后春笋。今天,网络学习资源库已经相当的丰富。然而许多资源库却没有得到充分的利用,特别是几何学习资源库! 笔者对几何学习资源库利用率低的现状进行了调查研究,发现几何学习资源库中的资源大部分是以bmp、jpg、gif等格式存储,这给使用者带来了以下两大难题:1、在几何学习资源库中很难搜索到符合要求的资源;2、找到的资源不能动态几何变化。 针对以上问题,本文提出解决问题的方案——开发一个具有图像识别功能的动态几何系统。 首先,本文从图像识别入手,分析和比较了多种识别算法——Hough变换识别算法以及Hough变换的改进算法、Freeman链码算法以及Freeman链码算法的改进算法和基于单义域的图像识别算法。根据比较的结果,选择了最优的识别算法——基于单义域的图像识别算法作为识别系统算法。并设计了系统的层次结构,为系统的各个部分设计了相应的类,利用图像识别算法实现了对bmp图片中的点、直线、圆等几何图形的识别。 其次,识别了图形元素之间的几何关系。在图像识别过程中,得到的图形元素都是相互独立,它们之间是没有几何关系。通过识别图形元素在原图中的空间位置关系,系统在识别出的图形元素之间建立了相应的几何关系。在本系统中,主要识别的几何关系有以下几种:线上的点、平行线、垂直、直线的相交、直线与圆相交、圆与圆相交、等腰三角形和等边三角形。 最后,实现了动态几何。通过设置图形元素之间的约束与被约束条件建立了动态几何。 With the development of computer science and information technology, the Web-Based Learning Resource Library /(WBLRL/) spring up like mushrooms. There are plenty of WBLRLs nowadays, but many of them are inefficient, people cannot make full use of them, especially for Geometry Learning Resource Library. The author noticed that most of the geometric figures saved in the resource library were in formats as bmp, jpg, png etc. This resulted in two big problems for the users: 1. It is very difficult to search the required information; 2. the geometric figures in the resource library cannot perform dynamic geometry transform. This paper put forward a solution to these problems—design a dynamic geometry system based on image recognition. Firstly, the paper discussed image recognition. From analyzing some recognition algorithm such as Hough transform algorithm and its improvements, Freeman chain code algorithm and its improvements, and image recognition algorithm using primitive region, the most efficient recognition algorithm was chosen—image recognition algorithm using primitive region. By designing the hierarchy of the system and the class of each part, the system has realized the recognition of points, lines, and circles in bmp picture. Secondly, the paper discussed the recognition of geometric relation between any two graphic elements. In the first stage of image recognition, all the graphic elements are independent. According to the positions of the graphic elements in the original picture, the system constructs the corresponding relationship between them. The main geometric relations recognized in the system are as follows: point on a line, parallel lines, perpendicular lines, lines intersection, intersection between line and circle, intersection between two circles, isosceles triangle and regular triangle. Finally, the paper realized the dynamic geometry part of the system. The main work in the realization of dynamic geometry was to construct the restricted relations of the graphic elements.

关 键 词: 图像识别 变换 单义域 最小二乘法 动态几何

分 类 号: [TP391.41]

领  域: [自动化与计算机技术] [自动化与计算机技术]

相关作者

作者 熊建平
作者 谢玉婷
作者 王广华
作者 江春燕
作者 何佳宁

相关机构对象

机构 中山大学
机构 广州大学数学与信息科学学院
机构 深圳职业技术学院
机构 暨南大学
机构 华南理工大学工商管理学院

相关领域作者

作者 李文姬
作者 邵慧君
作者 杜松华
作者 周国林
作者 邢弘昊