Research On Urban Vegetation Information Extraction And Analysis Based On High Resolution Satellite Image

City vegetation is an importance part of the city ecosystem which is the outcome of various factors coming from nature and human beings. The level of urban greenery is regarded as an important factor of the city's sustainable development. To realize the target of the city's sustainable development, we should synthesize the theories of several subjects to investigate the problems such as the reasonable scale of the gross green-land, the construction of the urban vegetation type, the distribute form and how to realize the coupling of green and the city's function and conformation in the limited city space.In recent years, high resolution satellite image get extensive valuing along with the development of the remote sensing technology. Carrying out the research of the city application of the very high resolution satellite image has possibility, superiority and practicability which is a strong support of the informationization, modernization and siciencenization of the city's decision-making and management. It is also an important base of the city's sustainable development. Getting the information of the type, distribution and structure of urban vegetation based on the application of high resolution satellite image and its information extraction methods, can offer a base for the planning of city Greenland, optimizing the vegetation structure and improving city environment quantity, which is very significant for the development of future city.Applying high resolution IKONOS image, this paper made a case study in XuanWu District, Nanjing City by introducing KPCA SAM classification method in extracting urban vegetation information and elementary establishing the remote sensing estimation model about urban "Green Quantity". The aim is increasing the accuracy of urban vegetation classification and caring out new idea of city "Green Quantity" investigation, which is hoping to providing scientific basis for estimating the benefit of urban ecosystem, reflecting the variation of city ecosystem and offering reference for the building of "Ecosystem Garden City" in Nanjing. The main contents and research achievements are as following:1. The extraction and rectification of shadow is a big problem in the pro-processing of high resolution satellite image. This paper adopts different method to extract and rectify mountain shadow and building shadow. For the building shadow in downtown area, adapting the Shadow Automated Extraction Method of IKONOS Images Based on Image Fusion to extract and the Lamber model to rectify; for the mountain shadow, directly adapting supervised classification method to extract and histogram matching method to rectify. The experiment shows the strategy can obtain an ideal result by adapting the strategy of separately extraction and rectification.2. Introducing KPCA SAM classification method in extracting urban vegetation information from high resolution image. The selection of kernel-function, including the selection of the function's parameter and the suitable number of training samples in the KPCA transform method is discussed. A KPCA SAM urban vegetation classification method is established and experiments shows a good result with gross accuracy of 80.6% involving all types and 91.7% considering first class types which show that this method can acquire more accurate data of urban vegetation and offer better foundation data for urban building.3. Elementary establishing the remote sensing estimation model about urban "Green Quantity" adapting methods of step regress and Neural Network based on the analysis of vegetation indices and texture, which carries out new idea of city "Green Quantity" investigation and reveals some meaning for referring. Experiment shows that remote sensing information can be used for as the foundation of estimation of urban "GreenQuality" and the method is effective. As far as the accuracy is concerned, the Neural Network model is higher than the step regress model and the former is better. 4. Ecological evaluation and landscape analysis is made using Landscape Ecology theory and var...