Nasuro K. Models and means of neural network classification of satellite data

Ph.D. thesis is devoted to improve the classification of satellite data by the development of effective methods, models, implementation of classification and construction of intelligent software for automatic satellite monitoring in real time. An intelligent approach to the classification of satellite data was improved by constructing modular neural network architecture. An object model for the problem of classification of satellite data and a UML profile were developed to represent the neural networks. An integrated model of neural network classification of the landscape according to remote sensing of the Earth was proposed and practically implemented. This model is based on the developed neural network method of satellite data classification.