Collaborative Space Object Detection with Multi-Satellite Viewpoints in LEO Constellations
Researchers have developed a deep learning framework to improve space object detection (SOD) by fusing data from multiple satellite viewpoints. Their experiments, using YOLO-based detectors, demonstrated that multi-view input significantly enhances detection accuracy, with one configuration boosting mAP50 by 36.3% and mAP50-95 by 46.5%. This multi-view fusion strategy offers a viable and effective approach for enhancing space situational awareness in increasingly congested low Earth orbit constellations. AI
IMPACT Enhances space situational awareness, crucial for managing LEO congestion and ensuring space safety.