Intelligent Interpretation And Change Analysis Of Remote Sensing Image On The South Bank Of Hangzhou Bay

Coastal zone is the area where the sea and land intersect.Affected by natural and human factors,the coastal zone resources are constantly degrading.The rapid and accurate monitoring of the dynamic changes of the coastal zone is of great practical significance for the real-time update of geographic information,resource investigation and scientific management.This article aims to study the rapid and accurate extraction of coastal zone information from coastal zone remote sensing images and analysis of their changes.Taking the South Bank of Hangzhou Bay as the research area,this paper studies the optimization of sample information and the technology of image intelligent classification,builds an intelligent hierarchical classification model,interprets the Landsat remote sensing images of 1990-2015 in the South Bank of Hangzhou Bay,and studies and analyzes the land change in the research area from the perspective of time and space.From the perspective of the study area of land change in many aspects of research and analysis.The main content of the paper is as follows:1.Aiming at the problem of sample selection,the spectral features,texture features and geometric features of the image are comprehensively analyzed,and a combination of multiple features are used to extract the remote sensing image samples on the south bank of Hangzhou Bay.The experimental results show that based on the spectral information and the spatial feature information of the surface features,the separability of the surface features can be improved,and the information of the surface features of the coastal zone can be obtained quickly and accurately,which can provide high-precision classification mapping.2.In view of the complex ground features in the study area and the mixed pixel phenomenon in the remote sensing image,algorithms such as decision tree,cluster segmentation and support vector machine are applied to the classification of coastal zone remote sensing images,a hierarchical classification model suitable for the land use of the coastal zone is constructed to achieve the orderly and accurate extraction of samples.Experiments show that the classification results of the hierarchical classification model are all up to 90%,which can be used for remote sensing fine mapping of coastal zone.3.Based on the land surface classification data from 1990 to 2015 in the South Bank of Hangzhou Bay,the temporal and spatial changes were analyzed,and the driving factors were analyzed qualitatively.The results showed that the area of forest land,fallow land and construction land increased most obviously.And human factors and social-economic factors are the main influencing factors affecting land use change.