Study And System Design On Land Use Classification Based On Remote Sensing

Global environmental change has become increasingly fierce.Land use change is one of the core contents of current global environmental change research.Timely and accurate access to land use information provides the basis and important basis for government depart-ments to carry out land use planning,land and resources surveys,basic geographic infor-mation updates,environmental resources protection,land dynamic monitoring,and analysis and decision making.Land use classification is one of its basic tasks.With the development of the world’s space technology,remote sensing technology has become more and more ma-ture.It has the advantages of macroscopic,rapid,and accurate acquisition of terrestrial image information at all spatial scales.It has become one of the most effective tools and techniques for land use classification.In this paper,the domestic high-resolution satellite remote sensing imagery is selected to conduct a preliminary study on land use classification in Jiaji Town,Qionghai City,Hainan Province.Firstly,PMS image data with high scores were preprocessed.In the pretreatment pro-cess,Gram-Schmidt Pan Sharpening and NNDiffuse Pan Sharpening algorithms were used to fuse PMS panchromatic images and multispectral images,respectively.Image,the results show that the image quality of the NNDiffuse Pan Sharpening algorithm is higher than that of the Gram-Schmidt Pan Sharpening fusion image;then the maximum likelihood method,the minimum distance method,the ISODATA method,the K-Means method,and the object-oriented method are used respectively.The land use of Jiaji Town was classified,and the classification results obtained by these classification methods were evaluated using the con-fusion matrix.The results showed that the land use classification was based on the PMS image data of Jiaji Town,and the support vector machine(SVM)was object-oriented.The classification method can effectively improve the accuracy of the classification results;sec-ondly,the space-time analysis of the land use status of Jiaji Town shows that the current situation of land use classification in Jiaji Town of 2016 is 29.81%of the cultivated land,34.67%of the forest land area,and the water area.The area accounts for 5.01%,the urban and rural residents’ land area accounts for 24.85%,and the unused land area accounts for 5.8%.65%;In the analysis of the land use status of Jiaji Town,a method based on multi-source remote sensing analysis was adopted.The remote sensing of Landsat TM/ETM+ in Jiaji Town of Liangjing in August 2002 and August 2009 was selected respectively.Based on the data,the land use status of Jiaji Town in 2002 and 2009 was analyzed,and the land use change situation in Jiaji Town from 2002 to 2009 and from 2009 to 2016 was obtained.Finally,in the way of ENVI+IDL,Use the development environment of ENVI5.3 and IDL8.5 to design the land use classification system.In the analysis of the land use status of Jiaji Town,a method based on multi-source remote sensing analysis was adopted.Landsat TM/ETM+ remote sensing data of Jiaji Town in 2009 was selected to analyze the land use of Jiaji Town in 2009.The situation has led to the change of land use in Jiaji Town between 2009 and 2016.This paper discusses the preprocessing of PMS imagery of high-score No.1,evaluates the accuracy of several commonly used land use methods,analyzes the spatial-temporal characteristics of land use in Jiaji Town,and conducts a preliminary design of the land use classification system.