AUTOMATION OF GEOINFORMATION DATA PROCESSING BY NEURAL NETWORK TECHNOLOGY

Abstract

Methodology. This paper presents an optimized method of geodata processing based on deep neural networks (GeoDNN+ 2.0), which offers an efficient and automated approach to geographic data analysis. The introductory section details the shortcomings of existing geoinformation processing systems, such as complexity of setup, limited resources and processing time, unreliable results, and the need for constant monitoring. It emphasizes the need for a new approach that addresses these shortcomings and opens up new perspectives for geoinformatics.

Findings. GeoDNN+ 2.0 is based on deep neural networks, which are used to automatically determine the characteristics of geographic objects without prior manual data processing. This avoids human errors and improves processing accuracy, while self-learning mechanisms ensure continuous quality improvement using new data. GeoDNN+ 2.0 demonstrates high efficiency in processing large volumes of geodata due to its optimized structure and parallel computing.

Another important advantage of GeoDNN+ 2.0 is its ability to effectively integrate with various geodata sources, which allows you to get the full amount of information and provides comprehensive analysis. This is especially useful in areas where a large amount of data from different sources requires in-depth analysis and convenient integration tools.

The overall goal of the article is to reveal the potential of GeoDNN+ 2.0 in solving various geoinformatics problems and show its advantages over existing systems. To achieve this goal, the authors present a detailed analysis of the GeoDNN+ 2.0 architecture and provide examples of its implementation on the example of classifying geographic objects in shapefile format. In general, the article demonstrates that GeoDNN+ 2.0 is a powerful and promising tool for modern geoinformatics that solves the shortcomings of existing systems and provides optimised opportunities for automated processing and analysis of geodata, taking into account the needs of the modern world.