Research On Classification And Thematic Mapping In Remote Sensing Extraction

With the development of remote sensing technology, classification of remote sensing is one of indispensable parts of land utilization, monitoring,etc,and accuracy of classification directly affects the practical value of remote sensing data.Thematic maps using remote sensing data has become an important method of thematic mapping. Remote sensing images has many advantages, such as up-to-date state,macroproperty, short mapping cycle and the advantages of three-dimensional coverage,etc,which has played an important role in land utilization information access.Remote sensing image has been important means of thematic information extraction and mapping.Thematic maps using remote sensing images not only can greatly improve the accuracy of mapping, but also can save a lot of manpower and time.This paper selects a region of Changdu as the experimental data,after the pre-processing of the images, using several methods based on land utilization Category to carry on classification. With several classification methods accuracy comparison,Chose a higher accuracy classification method and obtain a higher accuracy classification images,then carrying on remote sensing research and experiments of thematic mapping.The main research contents and results are as follows:(1)After the pre-processing of images,this paper uses non-supervised classification, maximum likelihood supervised classification and decision tree classification method based on vegetation indices to execute classificatioon of remote sensing image. The results show that decision tree classification based on vegetation indices have a higher classification accuracy.(2)Though field survey and visual interpretation, define the categories of classification by linking region photos to remote sensing image.(3)This Paper uses C5.0 decision tree classification algorithm based on vegetation index to construct decision tree.(4)This paper carrys out a experimental study of remote sensing vegetation mapping project based on results of classification and cuts classification of images by the statistical unit to count vegetation area,then making thematic maps in support of GIS software.