Automatic extraction of urban land-use information from remote sensing images based on improved NDBI method

With regard to low precision and objective deficiency in the extraction of urban land-use information from remote sensing images,the spectral characteristics of main land covers on Landsat ETM images were analyzed.An automatic extraction method of urban land-use information from remote sensing images based on the improved normalized difference built-up index(NDBI) was proposed.The proposed method was verified by use of the Landsat ETM images in Nanjing City.The results show that the improved NDBI method can effectively extract the urban land-use information.Compared with the conventional NDBI methods,the improved NDBI method can eliminate the effects of sparse vegetation on the extraction precision of urban land-use information so as to raise the precision of the urban land-use information.It is of advantages of accuracy and objectiveness.