纹理是机器视觉

论文字数:10894.页数:25

摘要

 纹理是机器视觉一个很重要的研究领域。从应用角度来讲，纹理广泛应用于自动检测、医学影像处理、文档处理和遥感领域。可见纹理识别已得到了广泛的应用，所以对纹理识别进行研究具有重大的意义。根据技术的发展情况以及实现方法的可行性方面来考虑，本课题采用了统计法中的灰度共生矩阵方法来实现。灰度共生矩阵方法，是建立在估计图像的二阶组合条件概率密度函数基础上的，首先通过灰度图计算出4个角度的灰度共生矩阵，然后计算出每个像元对在灰度共生矩阵中的概率，最后根据能量，熵，惯性矩，相关，局部平稳的计算公式计算出这5个特征值。如此做出来的识别系统，能够比较准确地计算出纹理图像的特征值，对纹理进行识别非常得有效，这也证明了灰度共生矩阵方法是一种很有效的识别纹理的方法。

关键字： 纹理， 纹理识别， 灰度共生矩阵， VC++

Abstract

 Texture is a very important research field of machine vision. From application perspective, texture is widely used in the field of automatic detection, medical image processing, and document processing and remote sensing. Texture recognition has been widely used, and the research of texture recognition is of great significance. According to the condition of the Technical development and the feasibility of the arithmetic, the subject used Gray co-occurrence matrix method which belongs to the statistical method to achieve. Gray co-occurrence matrix method is based on the second-order conditional probability density function of the estimated image. First, compute the gray co-occurrence matrix of 4 angles from the gray image. Then, compute the probability of every pair pixel in the gray co-occurrence matrix. Finally, compute the 5 Characteristic values according to the formula of energy, entropy, inertia matrix, correlation, and local balance. The recognition system done by this way can calculate the Characteristic value of texture image accurately, and is very effective to recognize texture. This also proves that the co-occurrence matrix method is a very effective method to recognize texture.

Keywords: texture, texture recognition， gray co-occurrence matrix, VC++

目录

摘要 I
Abstract II
第1章 绪论 1
1.1 数字图像处理概念以及发展历史 1
1.2 研究的目的、背景和意义 2
1.3 国内外文献综述 3
1.4 关于纹理的一些其他方面的认识 4
1.4.1 纹理模型分类 5
1.4.2 纹理分析的应用领域 6
1.5 文本组织 6
第2章 关于Microsoft Visual C++ 6．0介绍 7
2.1 Visual C++ 6.0简介 7
2.2 Visual C++ 6.0的特点 7
2.3 Visual C++ 6．0常用向导简介 8
2.4 MFC类库 9
第3章 纹理分析的概念以及算法 11
3.1 纹理的概念 11
3.2 纹理分析的几种方法 11
3.3 灰度共生矩阵具体算法分析 12
3.4 程序实现 15
3.5 结果分析 16
第4章 总结和展望 20
致谢 21
参考文献 22