- 上一篇：基于GIS的徐州市加油站空间布局应用研究
- 下一篇：基于GIS的徐州市公共停车场分布特征研究
摘要土地利用信息的遥感监测是土地资源合理利用、科学规划和有效管理的基础。这是一个较为复杂的流程，虽然国内外学者开展了大量的研究，但是仍然存在很多问题和挑战。主要表现为，自动化程度不高和自动分类精度有待提高。本文在对国内外遥感图像分类方法充分研究分析的基础上，以中等空间分辨率Landsat TM数据为主要数据源，选择并使用决策树分类模型，根据矿区的土地利用特性和光谱特征，构建自动分类算法决策树，完成对徐州大屯矿区的土地利用遥感图像自动分类。试验结果表明：基于决策树的土地利用分类能够较好地完成矿区的土地利用自动提取，该方法总体分类精度达到86.26%，Kappa 系数为0.78。此外，决策树算法具有可塑性好、易修改、方便移植等优点，以上算法可为同行专家在矿区土地利用自动分类提供借鉴。68932
本论文有图8幅，表3个，参考文献20篇。
毕业论文关键词：决策树 土地利用 遥感影像 自动分类
The Classification of Land Use in Xuzhou Mining Area Based on Decision Tree
Abstract
Remote sensing monitoring of land use information is the basis of rational use ,scientific planning and effective management for land resources.This is a complex process, although scholars conducted a lot of research, but there are still many problems and challenges.The main performance is the low-level automation and the automatic classification accuracy needs to be improved.In this paper,based on the analysis of the methods for remote sensing image classification, the Landsat TM data of medium spatial resolution is the main data source.According to the characteristics and spectral characteristics of land use in the mining area,select and use the decision tree classification model,construct classification algorithm decision tree,to complete the remote sensing image automatic classification of Xuzhou DaTun mining area. The results showed that,land use classification based on decision tree can complete the automatic extraction of land use in mining area.The overall classification accuracy of this method is 86.26%, and the Kappa coefficient is 0.78.In addition, the decision tree algorithm has the advantages of good plasticity, easy to modify and easy to transplant.The above algorithm can be used for reference in the automatic classification of land use in mining area.
Key Words: Decision tree Land use Remote Sensing Image automatic classification
目录
摘要 I
Abstract II
1 绪论 1
1.1 研究背景和意义 1
1.2 国内外研究现状 1
1.3 研究的主要内容 2
2 研究区域概况与数据预处理 3
2.1研究区域概况 3
2.2遥感数据源与预处理 4
3 基于决策树的土地利用分类 5
3.1 决策树原理 5
3.2 特征分析与选择 6
3.3 原始光谱特征提取与分析 8
3.4 归一化植被指数特征提取分析 8
3.5 决策树分类流程图 9
3.6 地物信息的获取 10
4 信息提取结果精度评价 11
5 结论与讨论