Search academic texts

Information × Registration Number 0216U003328, 0113U003176 , R & D reports Title The development of intelligent land cover end land use classification methods using heterogeneous data popup.stage_title Head Kussul Nataliya, Registration Date 26-01-2016 Organization Space Research Institute popup.description2 The main aim of the project is to develop intelligent methods of land cover classification with use of satellite and ground data. The methods of research are following: structural and functional analysis, statistical methods, neural networks, object-oriented programming. Within the project machine learning methods and remote sensing data that can be used for land cover classification were analyzed, the methods of land cover classification with use of satellite data time series were developed. For achieving project's goal classification algorithms were analyzed in case of incomplete data and their reliable operation on incomplete input data was ensured. Developed methods were validated on the test polygons over the territory of Ukraine. With use of developed methods and land cover classification algorithms based on time series of satellite data and restoration of missing information thematic land cover maps for Ukraine were built. Developed methods and algorithms satisfy the principles of deep learning because for solving this applied task it is necessary to use big amount of geospatial and satellite data ("Big data"). Product Description popup.authors Басараб Руслан Михайлович Колотій Андрій Всеволодович Костецький Олександр Михайлович Кравченко Олексій Миколайович Куссуль Наталія Миколаївна Олійник Тамара Віталіївна Тютюнник Лідія Іванівна Шелестов Андрій Юрійович Яйлимов Богдан Ялкапович Ящук Дарья Юріївна popup.nrat_date 2020-04-02 Close