Computer vision algorithms for aerospace applications

这份提案描述了由NSERC在瑞尔森大学航空航天工程系计算机视觉管理实验室(CVML)举行的持续研究。奥库内娃博士的研究兴趣在于计算机视觉在航空航天工业中的应用。在空间领域，应用包括行星探索、在轨交会和航天器检查、自动对接和卫星维修业务。奥库内娃博士提出的研究计划研究计划包括以下项目：1)用于航天器目标距离图像3D配准的特征选择。该项目包括：研究和开发航天器目标的智能特征选择程序，以获得最佳的配准精度；2)研究迭代位姿估计算法的初始猜测的性质，特别是位姿估计问题中的哪些参数影响初始猜测以及如何进行最优选择；3)航天器表面目标布置的优化策略。该项目的目标是估计当目标在航天器舱表面形成优化集合时可能的精度改进。

This proposal describes continuing research sponsored by NSERC held in the Computer Vision Management Laboratory (CVML) at the Department of Aerospace Engineering of Ryerson University. Dr. Okouneva's research interests lie in the area of applications of computer vision to the aerospace industry. In the space domain, applications include planetary exploration, on-orbit rendezvous and spacecraft inspection, automatic docking and satellite servicing operations. Dr. Okouneva's proposed research program research program includes the following projects: 1) Feature selection for 3D registration of range images of spacecraft objects. This project includes the research and development of an intelligent procedure of feature selection on spacecraft objects to achieve the best registration accuracy; 2) Research on properties of the initial guess for iterative pose estimation algorithms, in particular, what parameters of a pose estimation problem affect the initial guess and how to optimally choose it; 3) Optimal Strategy of Target Placement on Spacecraft Surfaces. The goal of this project is to estimate a possible accuracy improvement when targets form an optimized set on the surface of a spacecraft module.