Робота системи технічного зору по елементах поверхні космічного апарата в умовах її неповної видимості

The problem of tracking reference objects by a computer vision system designed to estimate the relative position and attitude of uncooperative spacecraft at short distances is considered. This vision system is an integral part of the onboard control system for autonomous rendezvous and docking in sp...

Повний опис

Збережено в:
Бібліографічні деталі
Дата:2023
Автори: Melnychuk, Serhii, Shevchenko, Volodymyr
Формат: Стаття
Мова:Ukrainian
Опубліковано: V.M. Glushkov Institute of Cybernetics of NAS of Ukraine 2023
Теми:
Онлайн доступ:https://jais.net.ua/index.php/files/article/view/78
Теги: Додати тег
Немає тегів, Будьте першим, хто поставить тег для цього запису!
Назва журналу:Problems of Control and Informatics

Репозитарії

Problems of Control and Informatics
Опис
Резюме:The problem of tracking reference objects by a computer vision system designed to estimate the relative position and attitude of uncooperative spacecraft at short distances is considered. This vision system is an integral part of the onboard control system for autonomous rendezvous and docking in space. The problem under consideration arises when determining the initial data for the well-known PnP problem, in which the position and orientation of a body in space is determined by the images of control points, obtained by a perspective projection camera. The position of the control points relative to the coordinate system of the target body must be known. Since the target body is an abstract uncooperative spacecraft, only obvious visible inhomogeneities of its surface can be used to determine the control points. Therefore, the problem arises — how to choose control points so that they are tied to the surface of the body, and could be found with high accuracy in images under different external shooting conditions. One of the possible solutions is to use for binding small-sized structural elements of an uncooperative spacecraft, which are most likely present on its surface. Then the control point is bound to the known shape of the surface of the corresponding element, and its position on the image is determined through pattern recognition. This article discusses the case of using computer vision with a monocular camera. We consider operation at close distances, when only a part of the docking side of the target spacecraft is in the camera's field of view. Elements placed on its surface are used as reference objects. A method is proposed for finding the positions of control points on images for a certain class of reference objects during periodic shooting.