Space Target Detection Simulation Dataset

This is a simulation dataset for space target detection in optical imagery. The annotations follow the YOLO format. Each sequence file includes an SNR value in the last line for SNR-aware evaluation. The dataset is under continuous development.

The dataset covers multiple imaging modes and target motion speeds, enabling evaluation under diverse sensing and dynamic conditions. It is designed for training and benchmarking deep learning-based space object detection methods, especially under low-SNR scenarios.