基于国产高分一号卫星影像的城市绿地可达性研究及景观格局分析

发布时间：2018-02-12 16:09

本文关键词： GF-1影像影像融合面向对象分类绿地可达性景观格局　出处：《东华理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：城市绿地是城市景观的重要组成部分,具有改善城市生态环境,为城市提供各种社会效益和服务等重要作用。开展城市绿地可达性研究和景观空间格局分析,并探讨绿地空间格局对可达性的影响,可以评价城市绿地布局的合理性和公平性,对提高城市绿地的服务水平,改善居民生活环境都具有重要意义。本文以GIS和遥感技术为支撑,以南昌市东湖区和西湖区为研究对象,利用研究区“高分一号”(GF-1)影像为主要数据源,在对原始GF-1影像进行预处理的基础上,探讨不同融合方法对GF-1影像的适用性差异;然后利用基于规则集的面向对象分类方法对研究区GF-1影像进行土地利用信息提取;在此基础上,利用费用加权距离法对研究区城市绿地可达性进行了分级,并对研究区绿地可达性空间分布和服务水平做出了评价;最后对研究区绿地景观进行了斑块构成分析、形状指数分析、斑块破碎化分析和总体格局分析,探索了研究区绿地景观空间格局对可达性的影响,并对研究区绿地规划和建设工作提出了优化建议。主要研究成果如下:(1)明确了不同融合方法对GF-1影像的适用性差异。对于研究区GF-1影像,Brovey融合法的融合效果比较差;HPF融合法的影像光谱畸变最小,光谱信息保持最好;PCA融合法在增强空间信息和保留光谱信息两个方面的能力都趋于中等;NNDiffuse融合法虽然在光谱信息的保持上不是最好,但其地物可分性最好,地物层次最多,影像包含的信息最丰富。(2)通过大量试验得出适用于研究区GF-1遥感影像及本文研究目的的影像最佳分割尺度和合并尺度;选取研究区大量的地物对象样本进行分析,建立了适用于研究区GF-1影像各地物类型的提取规则,从而实现了面向对象的研究区土地利用信息提取。(3)确定了研究区绿地可达性等级。两个研究区的绿地可达性分布都不均衡,但是整体可达性较好,研究区内的大多数居民都居住在绿地可达性为一级和二级的区域内,且两个研究区的绿地服务水平相差不大。(4)分析出了研究区绿地景观空间格局特征。两个研究区景观结构单一,景观多样性程度不高;绿地空间分布均匀性有待进一步优化,整体景观聚集度较高;绿地斑块形状受人为干预程度较大,破坏了自然生态的演替。(5)总结出了绿地景观空间格局对可达性的影响情况。总体而言,绿地斑块面积越大、斑块数量越多、景观空间格局分布越均衡,城市绿地可达性就越好。
[Abstract]:Urban green space is an important part of urban landscape, which can improve the urban ecological environment and provide various social benefits and services to the city. The research on the accessibility of urban green space and the analysis of landscape spatial pattern are carried out. The paper also discusses the influence of green space pattern on accessibility, which can evaluate the rationality and fairness of urban green space layout and improve the service level of urban green space. It is of great significance to improve the living environment of residents. This paper takes the GIS and remote sensing technology as the support, takes the Donghu District and the West Lake District of Nanchang City as the research objects, and uses the GF-1 image of the research area as the main data source. On the basis of preprocessing the original GF-1 image, the applicability of different fusion methods to the GF-1 image is discussed, and then the land use information is extracted from the GF-1 image of the study area by using the object oriented classification method based on the rule set. On this basis, the paper classifies the accessibility of urban green space by using the cost-weighted distance method, and evaluates the spatial distribution and service level of the green space accessibility in the study area; finally, the patch composition of the green space landscape in the study area is analyzed. Shape index analysis, patch fragmentation analysis and overall pattern analysis were carried out to explore the influence of green landscape spatial pattern on accessibility in the study area. The main research results are as follows: 1) the applicability of different fusion methods to GF-1 images is clarified. The fusion effect of GF-1 image Brovey fusion method is compared in the study area. The difference HPF fusion method has the lowest spectral distortion. The ability of PCA fusion method to enhance spatial information and preserve spectral information tends to be moderate, although the spectral information is not the best, but it has the best separability and the most ground level. The image contains the most abundant information. 2) through a large number of experiments, the best segmentation scale and merging scale for the GF-1 remote sensing image in the research area and the purpose of this study are obtained, and a large number of samples of ground objects are selected for analysis in the study area. The extraction rules applicable to the types of objects in the GF-1 image of the study area were established, and the object oriented land use information extraction in the study area was realized. The gradation of the greenbelt was determined. The distribution of the green land accessibility in the two study areas was not balanced. But the overall accessibility is good, and most of the residents in the study area live in areas where the greenbelt accessibility is grade one and two. And the difference of green space service level between the two study areas is not different. 4) the spatial pattern characteristics of green space in the study area are analyzed. The landscape structure of the two study areas is single, the degree of landscape diversity is not high, and the spatial distribution uniformity of green space needs to be further optimized. The overall landscape aggregation degree is relatively high, and the shape of green patch is affected by artificial intervention, which destroys the succession of natural ecology.) the influence of green space landscape spatial pattern on accessibility is summarized. In general, the larger the green patch area, the larger the green patch area. The more patches, the more balanced the distribution of landscape spatial pattern, the better the accessibility of urban green space.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P901;P237;TU985.2

【参考文献】