Comparative study of two detection methods based on land use data in Shenyang City
To investigate the effectiveness and applicability of maximum likelihood classification and object-oriented classification in land cover use detection,a comparative study was conducted on land use detection methods in the urban area of Shenyang City.Based on the Landsat remote sensing images of the urban area of Shenyang City in 1989,2000,2010 and 2016,the land cover/use classification of remote sensing images was divided into five types:water area,construction land,cultivated land,other vegetation,and other land.By comparing the overall accuracy and Kap-pa coefficient of the two experimental processes and results,it evaluated the strengths and weak-nesses of the two detection methods in practical applications.The overall accuracy of the pixel-based maximum likelihood method and object-oriented classification method exceeded 80%and 90%,respectively,with Kappa coefficients greater than 0.8 and greater than or close to 0.9,re-spectively.The object-oriented image classification method demonstrates high accuracy and ideal results,making it more suitable for the classification processing of high-resolution remote sensing images.
NETL
NSTL
万方数据
