| Найдено документов - 1 | Статьи из номера журнала: RUSSIAN AERONAUTICS. T. 69, № 1. - Moscow : Pleiades Publishing, Ltd., 2026. - Текст : электронный. | Версия для печати |
Сортировать по:
1. Статья из журнала
| Zvonarev, D. A. (Звонарев, Д. А.). Algorithm for target and attitude indicator recognition based on image preprocessing for lightweight neural networks / D. A. Zvonarev, T. Yu. Zhoraev. - Текст : электронный // RUSSIAN AERONAUTICS. - Moscow : Pleiades Publishing, Ltd, 2026. - T. 69, № 1. - pp. 97-110. - URL: https://elibrary.ru/item.asp?id=91680354 (дата обращения: 24.07.2026). - URL: https://link.springer.com/article/10.3103/S1068799826010095 (дата обращения: 24.07.2026). - Режим доступа: национальная подписка. | |
| Авторы: | Zvonarev, D. A., Zhoraev, T. Yu. |
| Ключевые слова: | Line detection, Image processing, Object detection, Iimage recognition, Object classification |
| Аннотация: | The problem of maintaining an unmanned aerial vehicle in space is solved by using landmarks and calculating the approximate speed using a pixel stream. The implementation of primary image processing algorithms using one-dimensional digital filters adapted for a digital educational environment is considered. To improve performance, methods of replacing high-order bits of operands with low-order ones, index arithmetic, and methods using the graphic object identifiers, as well as replacing nonlinear operations with binary ones, are used. Algorithms for extracting lines in images and trajectories based on lines forming open or closed contours are developed. An artificial neural network for recognizing arbitrary classes of polygons using artificial intelligence tools is synthesized |
| Поиск: | Источник |
| Ссылка на ресурс: | https://elibrary.ru/item.asp?id=91680354 |
| Электронный документ | Для просмотра необходимо войти в личный кабинет |