Research On Land Use Information Extraction With High-resolution Remote Sensed Image

In recent years,the spatial resolution of remote sensing images has become increasingly popular and data acquisition has become more convenient,making it possible to more clearly express the features,such as geometry,texture,etc.Therefore,it was widely used in geological,environmental,agriculture,forestry,hydrology,and meteorology fields.The traditional pixel-based classification technology has low precision in extracting high-resolution remote sensing information.Therefore,how to extract ground information on high-resolution remote sensing images effectively and accurately becomes an important research content in the field of remote sensing It is impossible to extract feature information more accurately on high-resolution remote sensing images.Object-oriented image classification technology has been widely used in high-resolution remote sensing images with higher classification accuracy.This paper,taking the extraction of land use for soil erosion monitoring in Ji County,Tianjin as an example,focuses on two points to improve the automation degree and classification accuracy of land use information,based on the domestic high-resolution GF-1image data and other ancillary data.The first is image segmentation,which is a complex and uniform distribution of multi-scale features for high-resolution images.It cannot describe its characteristics with a single scale.This paper uses the largest area method and image feature analysis method to determine the optimal scale from both quantitative and qualitative perspectives.The three hierarchical levels,Level 60,Level 85,and Level 120,are respectively determined as suitable extraction scales for the corresponding land use typ In recent years,the spatial resolution of remote sensing images has become increasingly high,making it possible to more clearly express the features,such as geometry,texture,structure,spectrum,etc.Therefore,it was widely used in geological,environmental,agriculture,forestry,hydrology,and meteorology fields.The traditional pixel-based classification technology has low precision in extracting high-resolution remote sensing information.Therefore,how to extract ground information on high-resolution remote sensing images effectively and accurately becomes an important research content in the field of remote sensing.Object-oriented image classification technology has been widely used in high-resolution remote sensing images with higher classification accuracy.