To obtain access to full text of journal and articles you must register!
- Article name
- RESULTS OF EXPERIMENTAL STUDIES OF AUTOMATIC RECOGNITION OF SPACE OBJECTS FROM SYNTHESIZED SPECIES DATA USING CONVOLUTIONAL NEURAL NETWORKS
- Authors
- Shavin A. S., , vka@mil.ru, Military Space Academy named after A. F. Mozayskij, St.-Petersburg, Russia
Khlebnikov S. G., , vka@mil.ru, Military Space Academy named after A. F. Mozayskij, St.-Petersburg, Russia
Kotyashov E. V., , vka@mil.ru, Military Space Academy named after A. F. Mozayskij, St. Petersburg, Russia
- Keywords
- automated recognition of space objects / convolutional neural networks / processing of species data / monitoring of space debris / optimization algorithms / synthesized data
- Year
- 2024 Issue 4 Pages 3 - 8
- Code EDN
- WQESEL
- Code DOI
- 10.52190/1729-6552_2024_4_3
- Abstract
- This article presents the results of research characterizing the possibility of using neuron networks for automatic recognition of space objects from synthesized species data. Based on the analysis of existing approaches to recognizing objects and obtaining data for training neural networks, it was found that a promising direction in the field of space situation analysis is the training of convolutional neural networks on synthesized data on space objects. Particular attention is paid to the problems associated with the effective use of neural network technologies, for which the issue of determining optimal parameters for improving the accuracy of convolutional neural networks has been investigated.
- Text
- BUY for read the full text of article
- Buy
- 500.00 rub
