IMPROVEMENT THE ACCURACY OF LAND COVER/LAND USE CLASSIFICATION USING FUZZY LOGIC AND EBBI INDEX FROM SENTINEL 2 AND LANDSAT 8 SATELLITE IMAGES

Abstract

In recent years, land cover/land use changes very quickly due to the impact of population growth and socio-economic development. Classification of land use/land cover from multispectral remote sensing data is often difficult and has a low accuracy. This paper presents the method of combination of Sentinel 2 and Landsat 8 satellite images for land cover/land use classification using fuzzy techniques. The extraction of bare land and built-up area can be done through the use of EBBI (Enhanced Built-up and Barreness Index) index. Middle infrared band (band 11), near infrared band (band 8) of Sentinel 2 image and thermal infrared band (band 10) of Landsat 8 image were used to calculate EBBI index. The obtained results showed that, method of land cover/land use classification based on combination of Sentinel 2 and Landsat 8 satellite images improve overall accuracy by about 5% compared with method using only Sentinel 2 data. The results obtained in the study can be used for the management, assessment and monitoring of the status and dynamics of land cover/land use.