Yailymov B. Automated information technology for land cover mapping based on satellite data fusion methods and models.

The dissertation is devoted to solving a problem of accuracy improvement of classification and timely evaluation of area classes of the land cover and support of high accuracy in mapping of the earth's surface on large areas by developing of the methods of fusion of different types of large volumes of geospatial data and creating of automated mapping technology. The method of classification of the land cover using satellite data of large volumes on the basis of the use of the ensemble of the neural networks was improved and the methods of fusion of different satellite data on pixels’ levels and decisions levels for mapping were implemented in automated information technology for Ukraine within a service-oriented approach. The method of mapping of the land cover was used in obtaining maps for the territory of Ukraine based on Landsat-4/5/7 for 2010, 2000 and 1990. The obtained maps made it possible to assess the general trends of various purpose of the land cover in Ukraine. The problem of the breach of crop rotation based on the developed technology was solved, the land cover maps were developed for the territory of Kyiv region for 2013-2015 years, the problem of assessing of the damages from drought was solved.