基于云理论的遥感影像分类方法研究

发布时间：2018-07-21 19:01

【摘要】：随着空间科技的发展，遥感技术已成为当前人类研究地球空间环境的一种有力技术手段。在遥感技术的研究中，无论是遥感信息提取、动态变化预测，还是专题地图制作和高分辨率遥感数据库的建立都离不开遥感影像分类，它是遥感技术中的热点研究内容。虽然遥感影像分类算法很多，但都不能同时解决遥感影像分类中的模糊不确定性和随机不确定性问题。而云理论能够将各概念中的模糊性和随机性有机的结合在一起，可以解决分类中的两种不确定性问题。因此，，本文将云理论引入遥感影像分类中，论文主要的研究内容如下： ○1深入研究了云定义、云的三个数字特征、基本云模型、云发生器及虚拟云等相关理论，并分析了云理论在解决分类过程中的模糊不确定性和随机不确定性问题方面的优势。 ○2按照云模型的差异，提出了三种基于云理论的遥感影像分类方法，即：基于改进正态云的遥感影像分类方法、基于浮动云的遥感影像分类方法和基于评语云与虚拟云的遥感影像分类方法。 ○3采用本文提出的三种基于云理论的分类方法进行了分类试验，并与传统的ISODATA聚类方法、最大似然分类法及模糊分类法进行了对比试验，分类结果表明基于云理论的遥感影像分类方法具有分类精度高，且计算过程简单等优点。
[Abstract]:With the development of space science and technology, remote sensing technology has become a powerful technical means to study the Earth's space environment. In the research of remote sensing technology, whether remote sensing information extraction, dynamic change prediction, thematic map making and the establishment of high-resolution remote sensing database are inseparable from remote sensing image classification, it is a hot research content in remote sensing technology. Although there are many remote sensing image classification algorithms, they can not solve the fuzzy uncertainty and random uncertainty in remote sensing image classification. Cloud theory can combine the fuzziness and randomness of various concepts together and solve the two kinds of uncertainty problems in classification. Therefore, this paper introduces cloud theory into remote sensing image classification. The main contents of this paper are as follows: 01 the definition of cloud, the three numerical features of cloud, the basic cloud model, cloud generator and virtual cloud theory are deeply studied. The advantages of cloud theory in solving the problem of fuzzy uncertainty and random uncertainty in classification are analyzed. 02 according to the difference of cloud model, three methods of remote sensing image classification based on cloud theory are proposed. That is, the classification method of remote sensing image based on improved normal cloud, The classification method of remote sensing image based on floating cloud and the classification method of remote sensing image based on comment cloud and virtual cloud. Compared with the traditional ISODATA clustering method, the maximum likelihood classification method and the fuzzy classification method, the classification results show that the remote sensing image classification method based on cloud theory has the advantages of high classification accuracy and simple calculation process.
【学位授予单位】：辽宁工程技术大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237

【参考文献】