南宁市西乡塘区城市绿地信息提取及景观格局分析

发布时间：2018-05-31 10:45

本文选题：高分二号 + 绿地信息　；参考：《广西大学》2017年硕士论文

【摘要】：城市绿地,作为生态建设的主要元素,能够改善生活环境和游憩条件,在经济效益、生态效益等方面具有不可替代的作用,城市绿地规划水平在不断进步的同时也相应逐渐暴露出诸多缺点和不足,过度的追求经济高速发展,甚至以破坏生态环境为代价,导致城市亚健康状态,因此运用最新高分二号感影像对城市绿地信息提取及分析对政府相关部门规划城市绿地有着很强的指导意义。2014年,随着国产高分二号遥感卫星的投入使用,在绿地信息提取方面有着极其重要的意义,有了信息化、现代化、科学化的技术支撑支持,城市绿地生态可持续研究更加完善。本文以南宁市西乡塘建成区为例,以2016年6月份的国产高分二号遥感影像作为主要信息源,对高分二号遥感影像在ENVI软件中进行辐射定标,RPC正射校正,FLAASH大气校正,图像配准,融合方法研究,根据国家绿地分类标准并结合研究区绿地实际状况,将城市绿地分为6种类型,并在Fragstats软件支持下进行城市绿地类型的景观格局分析等,为未来的南宁城市绿地景观规划提供合理依据。主要研究结果如下:(1)运用四种不同方式进行图像融合,通过主观定性和客观定量分析,得到NND融合法的保持光谱能力最强且空间纹理特征比较清晰。基于NND融合方法在融合速度上的研究发现,当多光谱的储存格式为BIP格式时融合速度快了将近3倍。(2)在绿地信息提取的过程中,运用面向对象的多尺度分割方法,经过实验样区的反复实验研究,分割阈值设置为35,合并阈值设置为75时效果最佳;面向对象提取植被信息规则应用:NDVI值为0.364,蓝色波段(band2)DN值小于1330,能够把绿地很好的提取出来。(3)景观指数分析得到,研究区城市绿地附属绿地的破碎化指数(FN2)为0.865,平均斑块面积(MPS)指数为0.042,说明附属绿地的破碎化程度高,绿地被分割严重。公共绿地MPS值相对较大,但总体面积较小且分布不均匀,街头游憩绿地少。(4)研究区景观多样化指数为1.531,景观均匀度指数为0.712,平均斑块分维数为1.385,整体破碎化程度较高,各种绿地类型面积分布相对均衡,但待开发区域以及城乡结合部绿地比较分散,整体性不强,未能形成连续的生态链条,生态功能和社会功能显著性低。
[Abstract]:Urban green space, as the main element of ecological construction, can improve the living environment and recreational conditions, and play an irreplaceable role in economic and ecological benefits. At the same time, the level of urban green space planning has gradually exposed many shortcomings and deficiencies, the excessive pursuit of rapid economic development, even at the cost of destroying the ecological environment, leading to sub-health state of the city. Therefore, the use of the latest high score 2 image has a strong guiding significance for the information extraction and analysis of urban green space. In 2014, with the use of the domestic high score 2 remote sensing satellite, it is of great significance for the government departments concerned to plan the urban green space. The extraction of green space information is of great significance. With the technical support of information, modernization and science, the study of ecological sustainability of urban green space is more perfect. In this paper, taking the XINGTang building area of Nanning city as an example, taking the domestic high score 2 remote sensing image in June 2016 as the main information source, the radiometric calibration of the Gaofen 2 remote sensing image in the ENVI software is carried out, and the FLAASH atmospheric correction and image registration are carried out. According to the national green space classification standard and the actual situation of green space in the study area, the urban green space is divided into six types, and the landscape pattern analysis of urban green space type is carried out under the support of Fragstats software. It provides a reasonable basis for the future planning of urban green space landscape in Nanning. The main results are as follows: (1) four different methods are used for image fusion. Through subjective qualitative and objective quantitative analysis, it is concluded that the NND fusion method has the strongest ability to maintain the spectrum and the spatial texture features are clear. Based on the research on the fusion speed of NND fusion method, it is found that the fusion speed is nearly three times faster when the multi-spectral storage format is BIP format.) in the process of green space information extraction, the object-oriented multi-scale segmentation method is used. After repeated experimental study in the experimental sample area, when the segmentation threshold is set to 35 and the combined threshold is set to 75, the effect is the best. Application of object oriented extraction of vegetation Information rules: the NDVI value is 0.364 and the blue band 2DN value is less than 1330, which can extract the green space very well. The fragmentation index (FN2) and average patch area (MPS) index of attached green space were 0.865 and 0.042 respectively, which indicated that the degree of fragmentation was high and the division of green space was serious. The MPS value of public green space is relatively large, but the total area is small and the distribution is uneven. The landscape diversity index, landscape evenness index and average patch fractal dimension of the study area are 1.531, 0.712, 1.385 and 1.385 respectively. The area distribution of various types of green space is relatively balanced, but the green space in the areas to be developed and the combination of urban and rural areas is relatively dispersed, the integrity is not strong, the continuous ecological chain has not been formed, and the ecological function and social function are low.
【学位授予单位】：广西大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S731.2

【参考文献】