遥感影像地表反射率及NDVI尺度效应研究
发布时间：2019-03-02 13:48
【摘要】：本文以广西省北海市的沙田半岛的部分区域为研究对象，以Geoeye-1遥感影像和Landsat7ETM+遥感影像作为数据源，经图像处理后，先从研究区整体上来分析0.5m、2m以及30m之间的地表反射率以及归一化植被指数(NDVI)的尺度效应，然后进行更为细致的研究实验，利用0.5m的Geoeye-1全色波段影像以及野外的观测数据在所研究的实验区域内选出七种具有代表性的地物样区，开展不同下垫面类型的地表反射率和归一化植被指数的尺度效应分析。本文具体的研究内容和取得的研究成果如下： （1）以往的研究尺度效应的方法大多是以一种遥感卫星影像作为数据源进行尺度效应的研究，往往在研究某种特定尺度的时候没有相应空间分辨率的卫星数据作为参考值，而本文则是采用两种卫星实际获取的遥感数据作为尺度效应研究中的参考值，并分为研究区域整体与不同下垫面类型的地表反射率与归一化植被指数尺度效应研究，，即从整体与相对微观的角度进行了全面的分析，这使得整个实验更具严密性和真实性。 （2）在数据准备阶段，为了获取0.5m的影像尺度，充分利用了Geoeye-1卫星数据中的0.5m全色影像与2m多光谱影像进行影像融合，在融合中运用了基于ENVI中的HSV变换、Brovey变换、Gram-Schmidt变换、主成分(PC)变换、colornormalized (CN)变换以及Pansharpening变换等六种变换方式，并同过定性分析与定量分析从中选出一幅影像质量最好的0.5m的融合影像作为研究中的实验数据，并分别对两种不同的传感器从光谱特性、成像几何特性以及时相特性进行了分析，并将两者的这三种特性进行了必要的归一化处理以降低实验研究所产生的差异。 （3）在实验结果与分析中，分别从所选择的研究区域的整体与不同下垫面类型两个不同的角度并选用五种评价指标进行了较为详细的研究与分析。在不同下垫面类型的研究中选取了研究区中的七种典型地物，包括红树林、桉树林、人工建筑、木薯地、裸地、耕地、水体等，发现随着尺度的增加，各种地物的细节信息逐渐减少，所选样本的灰度级逐渐变少，甚至有的信息消失殆尽。并利用空间异质指数来分析产生尺度效应的原因。
[Abstract]:In this paper, part of the Shatin Peninsula in Beihai City, Guangxi Province is taken as the research object, and Geoeye-1 remote sensing image and Landsat7ETM remote sensing image are used as the data source. After image processing, the whole area of the study area is analyzed 0.5m, The surface reflectivity between 2m and 30m and the scale effect of normalized vegetation index (NDVI) are studied in more detail. Using 0.5m Geoeye-1 panchromatic band images and field observation data, seven representative terrain samples were selected in the experimental area studied. The scale effect analysis of surface reflectivity and normalized vegetation index for different types of underlying surface was carried out. The specific research contents and achievements of this paper are as follows: (1) in the past, most of the methods used to study scale effects were based on a remote sensing satellite image as a data source to study the scale effects. Often when studying a particular scale, there is no corresponding spatial resolution satellite data as the reference value, but in this paper, the remote sensing data obtained by two kinds of satellites are used as the reference value in the scale effect research. The scale effects of land surface reflectivity and normalized vegetation index of the whole area and different types of underlying surface are studied, that is, the overall and relative microcosmic analysis is carried out, which makes the whole experiment more rigorous and authentic. (2) in the data preparation phase, in order to obtain 0.5m image scale, the 0.5m panchromatic image from Geoeye-1 satellite data and 2m multispectral image are fully utilized for image fusion, and the HSV transform based on ENVI is used in the fusion. Brovey transform, Gram-Schmidt transform, principal component (PC) transform, colornormalized (CN) transform and Pansharpening transform, etc. Through qualitative analysis and quantitative analysis, a 0.5m fusion image with the best image quality is selected as the experimental data in the study, and the spectral characteristics of two different sensors are analyzed respectively. The geometric characteristics of imaging are analyzed in time, and the three characteristics are normalized to reduce the differences in the experimental study. (3) in the experiment result and analysis, the research and analysis are carried out in detail from two different angles of the whole research area and the different type of underlying surface, and five evaluation indexes are selected to carry on the more detailed research and analysis. In the study of different types of underlying surfaces, seven typical features in the study area were selected, including mangroves, eucalyptus forests, artificial buildings, cassava fields, bare fields, cultivated land, water bodies, etc., and found that with the increase of scale, The detail information of all kinds of figures decreases gradually, the gray level of selected samples decreases gradually, and even some of the information vanishes. The spatial heterogeneity index is used to analyze the causes of scale effect.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237
本文编号：2433100
[Abstract]:In this paper, part of the Shatin Peninsula in Beihai City, Guangxi Province is taken as the research object, and Geoeye-1 remote sensing image and Landsat7ETM remote sensing image are used as the data source. After image processing, the whole area of the study area is analyzed 0.5m, The surface reflectivity between 2m and 30m and the scale effect of normalized vegetation index (NDVI) are studied in more detail. Using 0.5m Geoeye-1 panchromatic band images and field observation data, seven representative terrain samples were selected in the experimental area studied. The scale effect analysis of surface reflectivity and normalized vegetation index for different types of underlying surface was carried out. The specific research contents and achievements of this paper are as follows: (1) in the past, most of the methods used to study scale effects were based on a remote sensing satellite image as a data source to study the scale effects. Often when studying a particular scale, there is no corresponding spatial resolution satellite data as the reference value, but in this paper, the remote sensing data obtained by two kinds of satellites are used as the reference value in the scale effect research. The scale effects of land surface reflectivity and normalized vegetation index of the whole area and different types of underlying surface are studied, that is, the overall and relative microcosmic analysis is carried out, which makes the whole experiment more rigorous and authentic. (2) in the data preparation phase, in order to obtain 0.5m image scale, the 0.5m panchromatic image from Geoeye-1 satellite data and 2m multispectral image are fully utilized for image fusion, and the HSV transform based on ENVI is used in the fusion. Brovey transform, Gram-Schmidt transform, principal component (PC) transform, colornormalized (CN) transform and Pansharpening transform, etc. Through qualitative analysis and quantitative analysis, a 0.5m fusion image with the best image quality is selected as the experimental data in the study, and the spectral characteristics of two different sensors are analyzed respectively. The geometric characteristics of imaging are analyzed in time, and the three characteristics are normalized to reduce the differences in the experimental study. (3) in the experiment result and analysis, the research and analysis are carried out in detail from two different angles of the whole research area and the different type of underlying surface, and five evaluation indexes are selected to carry on the more detailed research and analysis. In the study of different types of underlying surfaces, seven typical features in the study area were selected, including mangroves, eucalyptus forests, artificial buildings, cassava fields, bare fields, cultivated land, water bodies, etc., and found that with the increase of scale, The detail information of all kinds of figures decreases gradually, the gray level of selected samples decreases gradually, and even some of the information vanishes. The spatial heterogeneity index is used to analyze the causes of scale effect.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：P237
【参考文献】
相关期刊论文 前10条
1 宋富强;康慕谊;杨朋;陈雅如;刘阳;邢开雄;;陕北地区GIMMS、SPOT--VGT和MODIS归一化植被指数的差异分析[J];北京林业大学学报;2010年04期
2 陈果;柳钦火;刘强;杜永明;;MODIS和降尺度TM数据反演叶面积指数相互验证中几何处理方法的研究[J];北京师范大学学报(自然科学版);2007年03期
3 张宏斌;杨桂霞;李刚;陈宝瑞;辛晓平;;基于MODIS NDVI和NOAA NDVI数据的空间尺度转换方法研究——以内蒙古草原区为例[J];草业科学;2009年10期
4 许辉熙;;遥感影像融合方法的精度评价[J];测绘与空间地理信息;2009年06期
5 李超;朱满;赵家平;;多源遥感影像融合效果的定量评价研究[J];测绘与空间地理信息;2010年03期
6 彭晓鹃,邓孺孺,刘小平;遥感尺度转换研究进展[J];地理与地理信息科学;2004年05期
7 陈崇成,黄方红,黄绚;自动散点控制回归技术在遥感数据辐射归一化中的应用[J];地球信息科学;2000年02期
8 苏理宏,李小文,黄裕霞;遥感尺度问题研究进展[J];地球科学进展;2001年04期
9 汪小钦;叶炜;江洪;;基于光谱归一化的阔叶林LAI遥感估算模型适用性分析[J];福州大学学报(自然科学版);2011年05期
10 余晓敏;陈云浩;;基于改进的自动散点控制回归算法的遥感影像相对辐射归一化[J];光学技术;2007年02期
相关博士学位论文 前1条
1 张雪红;基于高分辨率遥感的桉树林空间异质性与尺度效应研究[D];南京大学;2012年
相关硕士学位论文 前1条
1 胡婷;遥感图像典型地物特征提取的尺度效应研究[D];西北大学;2010年
本文编号：2433100
本文链接：https://www.wllwen.com/kejilunwen/dizhicehuilunwen/2433100.html
