A Pose Estimation Approach for Uncooperative Spacecraft Utilizing Convolutional Neural Network and Temporal Information

Accurate pose estimation is crucial for ensuring the safety and success of uncooperative spacecraft relative navigation missions. In recent years, convolutional neural networks (CNNs) have made significant progress in the field of spacecraft pose estimation due to their powerful feature learning capabilities. To further enhance estimation accuracy and robustness, this paper proposes a pose estimation algorithm for spacecraft based on temporal information and CNN. First, a novel neural network is introduced to achieve pose estimation for the spacecraft. Secondly, the temporal continuity of sequential images is employed for smooth optimization of the estimated results produced by the network, thereby reducing estimation errors. Finally, a dataset of sequential spacecraft images with a larger scale range is constructed on an advanced dataset and is used for experimental validation of the proposed algorithm. The results demonstrate that the proposed method effectively improves the accuracy of spacecraft pose estimation, accurately estimating the 6-DoF motion trajectory of spacecraft.