基于分形理论的NDVI连续空间尺度转换模型构建

发布时间：2018-05-29 14:15

本文选题：NDⅥ + 连续空间尺度转换　；参考：《南京大学》2013年博士论文

【摘要】：空间尺度转换是定量遥感重要而基础的问题之一,其中空间升尺度转换研究因其可用于解决如反演产品真实性检验等重要问题而受到广泛关注。目前的升尺度转换研究存在如下的问题：①基于离散的多传感器影像的反演量空间升尺度转换研究,所获得的尺度转换关系适用范围十分有限；②基于反演量“尺不变”的物理模型进行的反演量空间升尺度转换研究,囿于当前反演量“尺不变”物理模型的发展限制,其研究远不成熟。NDVI作为一种重要的反演量,其空间升尺度转换研究同样面临上述两方面的问题。针对已有研究存在的问题,论文提出基于分形理论通过构建NDVI连续空间尺度转换模型(a continuous spatial scaling model, CSSM)予以解决。此模型可揭示NDVI连续空间尺度转换规律,利用此模型有利于快速、有效地实现不同的大尺度(低空间分辨率)NDVI影像真实性检验。论文选取地类丰富、空间异质性显著的广西北海市沙田半岛作为研究区,综合利用Landsat ETM+、GEOEYE-1及HJ-1B CCD1等3种传感器数据,从以下三个方面对基于分形理论的NDVI连续尺度转换模型构建进行了系统研究：(1)基于分形理论的NDVI连续尺度转换模型构建；(2)遥感影像不同特性对模型构建的最合理尺度层级的影响分析；(3)NDVI连续尺度转换模型应用验证。论文主要研究内容和研究结论如下：1.基于分形理论构建了NDVI连续空间尺度转换模型,且论证了该模型适用于NDVI的真实性检验。以沙田半岛ETM+影像为例,基于分形理论构建得到该影像6个不同尺度层级Level (Level=33,66,100,150,200,250)下的NDVI连续尺度转换模型,基于模型构建的统计学一般性评价指标r、p、rlo、rup,以r最大、p
[Abstract]:Spatial scale conversion is one of the important and fundamental problems in quantitative remote sensing. The research of spatial scaling conversion is widely concerned because it can be used to solve such important problems as the verification of the authenticity of inversion products. At present, there are some problems in the research of scaling up scale conversion: 1. The scale conversion based on discrete multisensor image inversion is studied in space, and the scope of application of scale conversion relationship is very limited. (2) the spatial scaling transformation of inversion variables based on the "scale invariant" physical model is limited by the development of the current inversion "scale invariant" physical model, so its research is far from mature. NDVI is an important inversion quantity. The study of spatial scaling transformation also faces the above two problems. In view of the existing problems, this paper proposes that the fractal theory can be solved by constructing the NDVI continuous spatial scale conversion model (continuous spatial scaling model, CSSM). This model can reveal the law of NDVI continuous spatial scale conversion, and the model can be used to quickly and effectively test the authenticity of different large scale (low spatial resolution) NDVI images. In this paper, we select Shatin Peninsula of Beihai City, Guangxi Province, as the research area, which is rich in land species and significant spatial heterogeneity, and make comprehensive use of three kinds of sensor data, such as Landsat ETM GeOEYE-1 and HJ-1B CCD1, etc. A systematic study on the Construction of NDVI continuous scale conversion Model based on Fractal Theory from the following three aspects: 1) Fractal Theory based NDVI continuous scale conversion Model Construction of Guan2) remote Sensing Image with different characteristics is most reasonable for Model Construction Scale level impact analysis and application of NDVI continuous scale conversion model. The main contents and conclusions of this paper are as follows: 1. Based on fractal theory, the NDVI continuous spatial scale transformation model is constructed, and it is proved that the model is suitable for the authenticity test of NDVI. Taking the ETM image of the Shatin Peninsula as an example, based on fractal theory, the NDVI continuous scale conversion model with six different scale levels of Level level 3336100150200250) was constructed. The general statistical evaluation index rpplorupp was constructed based on the model, and the maximum r was used as the maximum value.
【学位授予单位】：南京大学
【学位级别】：博士
【学位授予年份】：2013
【分类号】：P237

【相似文献】