The high-precision attitude estimation technique for non-cooperative targets in space, based on monocular cameras, has important application value in missions such as space debris removal, autonomous rendezvous and docking, and on-orbit services. However, due to the inherent missing information problem of monocular vision systems and the high complexity of target geometry, existing monocular pose estimation methods find it difficult to realize an effective balance between accuracy and computational efficiency. Current solutions commonly adopt deep neural network architectures to improve estimation accuracy; but, this method is often accompanied by the problems of a dramatic expansion of the number of model parameters and a significant increase in computational complexity, which limits its deployment and real-time inference capabilities in real spatial tasks. To address the above problems, this paper proposes a spacecraft pose estimation network, called Balance-URSONet, which weighs the trade-off between accuracy and the number of parameters, and makes the pose estimation model have a stronger feature extraction capability by innovatively using RepVGG as the feature extraction network. In order to effectively improve the performance and inference speed of the model, this paper proposes the feature excitation unit (FEU), which is able to flexibly adjust the feature representation of the network and thus optimize the utilization efficiency of spatial and channel information. The experimental results show that the Balance-URSONet proposed in this paper has excellent performance in the spacecraft pose estimation task, with an ESA score of 0.13 and a parameter count 13 times lower than that of URSONet.
With the deep integration of Internet of Vehicles (IoV) and edge computing technologies, the spatiotemporal dynamics, burstiness, and load fluctuations of user requests pose severe challenges to microservices auto-scaling. Existing static or periodic resource adjustment strategies struggle to adapt to IoV edge environments and often neglect service dependencies and multi-objective optimization synergy, failing to fully utilize implicit regularities like the symmetry in spatiotemporal patterns. This paper proposes a dual-phase dynamic scaling mechanism: for long-term scaling, the Spatio-Temporal Graph Transformer (STGT) is employed to predict traffic flow by capturing correlations in spatial–temporal distributions of vehicle movements, and the improved Multi-objective Graph-based Proximal Policy Optimization (MGPPO) algorithm is applied for proactive resource optimization, balancing trade-offs among conflicting objectives. For short-term bursts, the Fast Load-Aware Auto-Scaling algorithm (FLA) enables rapid instance adjustment based on the M/M/S queuing model, maintaining balanced load distribution across edge nodes—a feature that aligns with the principle of symmetry in system design. The model comprehensively considers request latency, resource consumption, and load balancing, using a multi-objective reward function to guide optimal strategies. Experiments show that STGT significantly improves prediction accuracy, while the combination of MGPPO and FLA reduces request latency and enhances resource utilization stability, validating its effectiveness in dynamic IoV environments.
The predecessor of China Optics was China Optics and Applied Optics Digest, which was founded in 1985. At that time, it was the only retrieval journal in the field of optics in China. At the end of 2008, China Optics and Applied Optics Digest was renamed 《中国光学》的前身是1985年创办的《中国光学与应用光学文摘》,当时是国内光学领域唯一的检索刊物。2008年底《中国光学与应用光学文摘》转型更名为《中国光学与应用光学》期刊,刊载内容突破了文摘的限制,正式成为学术刊物。
1 September 2026 in Journal of Physics: Conference Series
Event cameras offer microsecond temporal resolution and high dynamic range, making them attractive for high-rate spacecraft attitude determination. However, existing event-based star-tracking methods usually rely on pre-calibrated optical parameters and cannot compensate for focal-length changes during operation. This paper presents STARE, a self-calibrating event-based star-tracking pipeline that identifies stars from raw event streams while recovering the effective focal length. STARE combines polarity-aware activity-adaptive spatiotemporal filtering, single-pass incremental clustering, and a golden-section multi-scale focal-length search for catalogue matching. In 13 laboratory recordings from a dynamic star simulator, comprising 2500 time windows acquired with an EVK4-HD event camera, STARE achieved 84.9% identification success with a mean reprojection error of 1.01 px. It recovered focal length under simulated deviations of up to ±20%, and tracking mode required approximately 1 ms per window, supporting 50 Hz attitude output. These results suggest that event-based sensing can provide calibration-tolerant attitude information for agile spacecraft.