Adopting Gram-Schmidt and Brovey Methods for Estimating Land Use and Land Cover Using Remote Sensing and Satellite Images

Abstract

The production of Land Use and Land Cover thematic maps using remote sensing data is one of the things that must be dealt with carefully to obtain accurate results, data is obtained from sensors of different characteristics. It is not possible to obtain high spatial and spectral accuracy in one image, so we used a fusion image (multispectral image with a low spatial resolution with a panchromatic image with high spatial resolution), which achieved high efficiency in improving the methods of producing Land Use and Land Cover maps. In this study, we used Landsat-8 multispectral and panchromatic images. The study aims to investigate the effectiveness of panchromatic images in improving the methods of producing Land Use and Land Cover maps for the city of Karbala, Iraq. The Support Vector Machine was used to classify the fusion images using the Brovey method and Gram-Schmidt sharpening algorithms. The appropriate methodology for producing Land Use and Land Cover maps was suggested by comparing classifying results and the classification accuracy was evaluated through the confusion matrix. Where the results showed that the method of classifying the fused image by Gram-Schmidt and classified by Support Vector Machine is the best way to produce Land use and Land cover maps for the study area and achieved the highest results for overall accuracy and kappa coefficient of 97.81% and 0.95, respectively.