<![CDATA[Land coverage of an island can be determined based on a multispectral image analysis. This research was carried out in Kei Kecil Island, Southeast Maluku Regency. The research aimed to determine land cover based on multispectral analysis of Landsat 8 (OLI) record on 27 November 2019. This research was carried out through several stages, namely pre-processing of image data (radiometric correction, correction geometric and image cutting), digital analysis of Landsat Image (Image Processing) and Accuracy Test. The classification method used was the Maximum Likelihood (MCL) by considering the prior probability factor, namely the chance of a pixel to be explained into a certain class. The results of Landsat 8 (OLI) image classification showed that there were 7 classes of land cover, with the coverage area of each land cover: settlements 34.73%, secondary forests 10.54%, water bodies 0.05%, shrubs 34.77%, mixed gardens 14.57%, open land 1.91%, and cloud 3.43%. The land cover of the multispectral image of 543 was dominated by shrubs, which was 34.8%, and the smallest was water body, which was 0.1%. In the multispectral image of 654, settlements dominated the land cover of the research area, which was 31.5% and the narrowest was open land, which was 0.9%. The accuracy was shown with an overall accuracy value of 88% and a Kappa score of 0.85%. This showed that the level of accuracy of classification results obtained through Landsat 8 multispectral image analysis (OLI) in 2019 had a very high level of accuracy (very good). These results met the requirements applied by USGS (United States Geological Survey).]]>
Copyrights © 2022
