Search academic texts

Information × Registration Number 0218U000415, 0117U007373 , R & D reports Title Investigation of the principles of formation of information products of international programs. Collection of available terrestrial and satellite data for the territory of Ukraine. popup.stage_title Head Kussul Nataliya, Registration Date 29-01-2018 Organization Space Research Institute popup.description2 The object of research is satellite images of high spatial resolution and methods of their processing. The purpose of the work is to explore and improve the models and methods of in-depth training for geospatial analysis in the interests of agriculture. Method of research - machine learning, geospatial analysis, object-oriented design, structural functional analysis, statistical methods, neural networks. With the advent of free access to large volumes of satellite data, the development of methods of machine learning based on geospatial data, in particular satellite, is becoming more and more relevant. The paper considers the basic methods of machine learning and analyzes the peculiarities and results of their application to the tasks of classification of terrestrial cover by satellite data of high distinction. Particular attention is paid to deep architectures, in particular, convergent neural networks, which today are the most powerful and precise method for recognizing visual images. The main advantages of the methods of deep studying over the traditional approaches to the classification tasks that have been used over the last decades and based on expert knowledge to distinguish the features from the input data are determined. The main principles of the formation of information products of international programs were investigated and a collection of terrestrial and satellite data for the territory of Ukraine was conducted. Product Description popup.authors Колотій Андрій Всеволодович Куссуль Наталія Миколаївна Яйлимов Богдан Ялкапович popup.nrat_date 2020-04-02 Close