Change Detection Method Of Geographic Situation Data Based On Heterogeneity

Geographical situation data is the synthesis and integration of various geographical information,reflecting the status of national natural resources and national ecological environment,and become an important basis for the formulation of national and regional development strategies and development planning.Able to express the surface characteristics and surface changes,through some change detection methods can detect the geographical situation data of two periods and get the results.This paper mainly studies the method of geographic situation data change detection,first introduces the change detection theory,data source,basic unit and change detection general process,the data source and basic unit of change detection is very important research object in the change detection process.The general process of change detection is divided into data preprocessing,geographical feature extraction,change detection and accuracy evaluation analysis.The data source used in this paper is the vector diagram of the old period and the remote sensing image of the new period,the basic unit of change detection is like spot.The change detection method is based on the geographic situation data change detection method based on heterogeneity,which in order to achieve the automatic change detection of vector graph and remote sensing image,a vector map and remote sensing image change detection method based on image banlei.Using the vector graph as constraints,the image segmentation of remote sensing image,the image segmentation of vector image constrained by the mark watershed algorithm,the traditional marker watershed algorithm is used to achieve image segmentation under the constraints of watershed After extracting the histogram of the image spot as the feature of image disk,the feature distance of pixels is converted to the distance of histogram,and the histogram distance is calculated by the algorithm of g statistics,and the heterogeneity between the image disk and the old period is constructed Using Otsu to obtain the heterogeneity threshold of each category,compare the category of pixel Spot with the category of belongs to heterogeneity,and realize the change / unchanged discrimination of image disk.Using the c + + language,based on the gdal(geospatial data abstraction)kaiyuanku and arcengine platform,the geographical situation normalized monitoring data change detection system.The systemimplements the functions of raster and vector data loading and overlay display,constrained image segmentation,pixel feature extraction,image banyi construction,and heterogeneity threshold acquisition.In order to verify the effectiveness of this method,in order to verify the effectiveness of this method,the accuracy of this method was improved by 0.08,the false positive rate decreased by 0.27 and 0.12 respectively.