Font Size: a A A

The Study Of Extracting The Rural Residential Area Based On High Resolution Image

Posted on:2018-06-18Degree:MasterType:Thesis
Country:ChinaCandidate:S Q XieFull Text:PDF
GTID:2310330515480126Subject:Cartography and Geographic Information System
Abstract/Summary:PDF Full Text Request
Today is the age of 3S integration development,and remote sensing is the basis of 3S,the launching of Gaofen-1,Gaofen-2 and Gaofen-3 satellites and product high resolution remote sensing image,which promoted the progress of The Times.In recent years,data update speed is accelerated,the spatial resolution of satellite remote sensing image has reached the level meters,highly increasing the application of remote sensing image,and the traditional method of visual interpretation thematic information,or extraction accuracy is not high automatic,semi-automatic extraction method,already could not satisfy the application requirements,and in each of the automatic extraction method can be applied to the extraction of the rural residential areas,can be applied to extract the rural residential areas of southern hillside area,little.therefore,high precision,streamline to extract hilly area of rural residential areas is the key to solve this problem.This Paper making Qiulin town as an example in Sichuan province Santai,bases on the Gaofen-2 and Gaofen-1 image,using ENVI software for image processing,to improve the effect of data display,reduce noise and improve the identification,gaining analysis again pixel spectral information and spatial shape features,texture information,spatial relationships,terrain factors,such as application of supervised classification and unsupervised classification,decision tree classification based on expert knowledge,based on object-oriented rules and sample information extraction and other methods to experimental research,implement target of feature extraction of rural building of residential areas.Then use the extraction results are internal tree,space analysis of the rural residential areas,to spatial analysis process,so as to realize the rural residential areas extraction of target feature,with visual interpreting of residential status vector superposition to evaluate the extraction accuracy.The main results of research are as follows:(1)This paper establishes a model for the extraction of rural residential buildings based on gaofen-2 data and the rule oriented feature extraction method,and the model of the extraction the urban residential area based on gaofen-1 data.(2)Object-oriented feature extraction method based on rules can be set through the spectrum and texture feature extracting rules are established,such as spatial feature information and attributes,the result also can more accurately extract the location of the rural residential areas,and its location accuracy above 85%.(3)Establish precision evaluation system,including the location accuracy,precision and range of comprehensive precision.(4)The study founds the distribution of villages and the size of villages is small,therefore,scientific and reasonable layout and scale of countryside are important.This article innovation points are as follows:(1)The first one uses Gaofen-2 image to extract was studied for the hilly area of rural residential areas and combine remote sensing processing software and GIS analysis software to extract the rural residential areas.(2)When researching characteristic information,gain the spectral characteristics,texture feature and shape feature space analysis,also analysis the terrain factors and spatial topological relations.With high revolution of remote sensing data,based on the research of high revolution remote sensing image and the study of rural residential areas,in the monitoring,geographical conditions census,city planning,housing construction,modern agriculture,disaster evaluation,environmental change predictions of application plays an important role,provide decision-making support for the sustainable development of urban and rural areas.
Keywords/Search Tags:high resolution remote sensing, the Rural Residential Area, ENVI, extraction, model of methods
PDF Full Text Request
Related items