A Neural Network Development for Multispectral Images Recognition

A three-layer neural network has been developed to recognise multispectral images using Python and the TensorFlow library. The constructed neural network will provide the class of the aircraft object according to the input data. Also, in this scientific article, the analytical construction of a neural network and a choice of the best activators and optimisers are considered. Finally, experiments with the developed neural network based on various input data are carried out.