Features of clustering the multispectral satellite Landsat images

Abstract: In the article, features of thematic processing of Earth remote sensing data obtained in different ranges of the optical spectrum using the Landsat-8 instrument are studied. The clustering of multispectral survey data, based on measurements in the first channels of a satellite instrument, with a resolution of 30 m, does not allow the high-precision determination of surface types due to heterogeneity of their mixing. The study of these aspects is the subject of this article. The results of real data processing of satellite measurements are presented.

Keywords: remote sensing, satellite image processing, multispectral, clustering

For citation:
Kataev M. Yu., Bogomolov A. V. Features of clustering the multispectral satellite Landsat images. Doklady Tomskogo gosudarstvennogo universiteta sistem upravleniya i radioelektroniki, 2018, vol. 21, no. 2, pp. 54–59. DOI: 10.21293/1818-0442-2018-21-2-54-59