ACCELERATION OF THE CLASSIFICATION OF SATELLITE IMAGES BY OPTIMAL IMAGE SEGMENTATION USING CONVOLUTIONAL NEURAL NETWORKS

The article discusses the methods of accelerating the satellite images classification. The main attention is paid to the problem of optimal separation, which is to minimize the loss of information in the image, as well as time and computing resources. Methods of pre -processing data and methods of combining classification results are presented. In conclusion we emphasize the importance of creating self -learning models that can automatically adapt the breakdown to specific data characteristics.