Dynamic Monitoring And Automatic Extraction Of Land Use Change Information With Supervised Classification

Land-use changes survey is the basis of the most important aspects in the annual statisticsand management of the land resources. Using remote sensing dynamic monitoring technology,not only reduces data preparation work of the early stage, but also reflect the field current landuse change objectively.But now, the main method is contrast the before and after phaseimaging through artificial to discovery variation pattern which monitoring results not only arethe subjective influence, and the data processing efficiency is not high. Based on the remotesensing dynamic monitoring study, using computer automatic classification technology,finding the change region, extracting question pattern, more rapid to provide basic data forinvestigation.This paper takes Shanxi province Baota District of Yanan city as a study area. By Summaryand comparison the remote sensing monitoring methods and analysis characteristics of image datain the experimentation area, use of the present land use remote sensing monitoring based on thesupervised classification of automatic discovery of change information method and extraction thechange information. This paper studies the three kinds of supervised classification methods,through the two phase remote sensing image change detection by land use type change areas,choose the best monitoring data. In order to better use GIS method for the analysis of district2011annual land use change dynamic research, it is necessary to monitor data processing two values,extracting change pattern and boundary vector data, and the vector data smoothing, extracting andprocessing.Through the experiment results and accuracy analysis conclusion: Based on the samecategory template, the most suitable supervised classification for the experimentation area isthe maximum likelihood method, the data processing has certain feasibility.