On October 2-6, 2023, Academician Ali Abbasov, General Director of the Institute of Control Systems (ICS) of the Ministry of Science and Education of the Republic of Azerbaijan, Associate Professor Adalat Pashayev, Head of the Laboratory of Modeling and Control of Continuous Processes, and Senior Researcher Elkhan Sabziev, DSc in Engineering, presented a paper at the 74th International Astronautical Congress (IAC). In the paper "On the object recognition in aerospace images", a recognition algorithm based on the properties of aerial images or objects in them that remain unchanged regardless of their orientation was developed.

 

The main idea of the method is that static moment properties of two-dimensional shapes can be applied to the images represented in the images. These properties include the "center of gravity" and "principal axes of inertia" of geometric figures. These properties do not depend on the orientation of the object seen from space, but are determined only with respect to its own shape. When the angle of image capturing changes little, these properties also change little. Calculation of the corresponding characteristics of objects in different aerospace surveys makes it possible to determine their orientation, thereby bringing them to a common standard and performing identification.

 

Based on the investigated approach, the paper also proposes the structure of an appropriate neural network and a learning mechanism for recognizing objects in aerospace images.

 

More than 5,000 representatives of space agencies, government agencies, investors and private space companies from 101 countries, including about 50 astronauts, took part in the congress in Azerbaijan, which serves the development and modernization of the space industry and advanced technologies. 150 companies were represented in the exhibition pavilion of the congress. The participating agencies and companies included space agencies of the USA, Russia, Türkiye, Japan, Italy, France, India and other countries.