基于蚁群算法和粗糙集方法的聚类分析研究-计算机应用技术专业论文.docxVIP

州llmilMMW刷刷 Tl?49841 摘要 摘 要 图形聚类分析是图像处理研究领域的热点之一。它的应用涉及到国民经济 的许多领域，目前已经有许多学者进行了大量的研究。由于将聚类分析应用到 图像模式识别可以对统…图像分类识别理论体系起到重要的作用，因此，本文 基于图形聚类分析的理论研究有着重要的意义。其主要工作包括以下几个方面: 首先，简要介绍了粗糙集理论和蚁群优化算法的相关知识，包括信息表达 系统，上近似和下近似，属性约简与核，属性依赖度和重要度等概念和对蚁群 算法的理论知识的总结概述。 其次，对图像分割与特征提取的方法进行了系统的综述。图像分割和测 是聚类分析的基础，图像分割将图像分割成一些有意义的区域，如何选取阙值 是图像能否正确分制的关键:此外，本文还对目前主要的聚类方法进行了深入 的研究与分析。 最后，本文提出了一种摹于相糙集和蚁群优化方法的聚类分析方法。该方 法充分利用蚁群算法的哥优能力，实现图像的识别。首先对所有样本进行编号， 然后提取数乎特征，再用蚁群算法进行数字聚类，最后用基于粗糙集理论的Jf 城关系进行聚类结果的评价。实验结果表明，所提出的方法是布效的、可行的。 关键词:蚁群算法:粗糙集;特征提取:聚类分析:模式识别 11 11 ABSTRAC 丁 ABSTRACT Graph clustering analysis is one of hot researches in the image processing field. It has been involved into many application fields of the national economy. the application of the national economy in many fields Nowadays ，a great deal of the researches have been done by many scholars. Because the application of cluster analysis in image pattem recognition plays an important role in the unity of image and recognition theory system，therefore，the study of graph clustering analysis theory is of great significance .This article mainly includes the several aspects as follows: Firstly，the thesis gives a brief introduction to the theories and methods of rough and ant colony optimization algorithm，which include information expression system， upper approximation and Iower approximation ，attribute reduction and core，attribute dependency，and the concepts of importance and the summary about the ant colony algorithm knowledge，etc. Secondly，the image segmentation and feature extraction method has been systematically summarized. Image segmentation and me削rement is the basis clustering analysis . The image is segmented with some recognized regions by the image segmentation. how to select the threshold is the key of image segmentation. In addition，the methods of clustering analysis have been studied and in the article. Finally，this paper puts forward a clustering analysis method based on rough set theory and optimization colony algor