基于支持向量机及MODIS数据的南昌市城市空间格局演变研究

发布时间：2018-05-10 20:09

本文选题：支持向量机 + MODIS　；参考：《东华理工大学》2016年硕士论文

【摘要】：随着经济的迅速发展及人口急剧地向城市聚集,南昌市城市化进程不断加快。城市化进程加快不仅表现在城市人口的增加、非农业活动比例的增加,城市建设用地扩张也是一个重要的衡量指标。及时掌握城市空间扩展信息,对国家、当地政府进行城市规划与建设、土地利用管理及发展方向决策都具有重要的意义。本文以城市化快速发展的南昌市为研究区,利用SVM、RS及GIS理论与方法进行南昌市城市空间格局演变研究。在整个研究过程中所做工作如下:(1)将支持向量机与传统的分类方法进行对比分析,结果发现利用支持向量机方法分类优于传统的分类方法;(2)分析单时相MODIS反射率数据、单时相加入指数信息、多时相反射率数据及多时相加入指数信息的不同特征组合方式对分类精度产生的影响,通过对分类精度的比较,决定选择多时相加入指数信息的分类特征参与城市空间格局演变研究;(3)利用支持向量机分类方法,以多时相加入指数信息的分类特征组合作为数据源,对南昌市2004、2009、2014年的遥感影像数据进行分类提取,获取南昌市十年间建设用地变化情况。同时利用Arc GIS10.0叠加分析功能对分类转矢量的建设用地变化情况进行分析,研究城市建成区空间格局演变轨迹;(4)从定性、定量两个方面分析南昌市城市空间格局演变驱动力。查找南昌市人口、生产总值、实际利用外资等数据,利用统计分析软件SPSS19.0构建主成分综合得分模型。依据该模型划分南昌市各行政区的空间格局演变强度,分析南昌市空间格局演变的趋势。空间格局演变研究可为掌握城市建设用地利用情况、控制城市建设用地范围以及判断城市总体发展方向等工作提供辅助决策信息。
[Abstract]:With the rapid development of economy and the rapid accumulation of population to the city, the urbanization process of Nanchang is accelerating. The acceleration of urbanization is not only reflected in the increase of urban population and the proportion of non-agricultural activities, but also the expansion of urban construction land is an important measure. It is of great significance for the state and local government to plan and construct the city, manage the land use and make the decision of the direction of development. In this paper, the urban spatial pattern evolution of Nanchang city is studied by using the theory and method of SVMRS and GIS, taking Nanchang city as the research area. The results show that the SVM method is superior to the traditional classification method in analyzing the single phase MODIS reflectivity data. The effects of different feature combinations of single-phase index information, multi-phase reflectivity data and multi-temporal phase index information on classification accuracy are compared. It is decided that the classification features of multi-temporal and exponential information should be selected to participate in the evolution of urban spatial pattern. (3) using support vector machine (SVM) classification method, the classification feature combination of multi-temporal adding exponential information is used as the data source. The remote sensing image data of Nanchang city in 2004 / 2009 and 2014 were extracted by classification and the change of construction land in Nanchang city during the past ten years was obtained. At the same time, by using the function of Arc GIS10.0 superposition analysis, the paper analyzes the change of construction land of classified transformation vector, and studies the evolution track of spatial pattern in urban built-up areas.) it analyzes the driving force of urban spatial pattern evolution in Nanchang from two aspects: qualitative and quantitative. This paper looks up the data of population, GDP and actual utilization of foreign capital in Nanchang city, and constructs a comprehensive score model of principal components by using statistical analysis software SPSS19.0. According to this model, the spatial pattern evolvement intensity of Nanchang city is divided, and the trend of spatial pattern evolution in Nanchang city is analyzed. The study of spatial pattern evolution can provide auxiliary decision information for mastering the utilization of urban construction land, controlling the scope of urban construction land and judging the overall development direction of the city.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：P208;TU984.113

【相似文献】