The SPARK 2022 dataset contains two streams of data.

Stream 1
Spacecraft Detection involves localising and classifying the object by estimating its bounding box and label. Stream-1 dataset is a unique space multi-modal annotated image dataset containing a total of ~110k RGB images of 11 object classes (10 spacecraft and one class of space debris). The data have been generated under a realistic space simulation environment, with a large diversity in sensing conditions, including extreme and challenging ones for different orbital scenarios, background noise, low signal-to-noise ratio (SNR), and high image contrast that defines actual space imagery.

Stream 2
Spacecraft Trajectory Estimation focuses on Leveraging the knowledge of temporal data to estimate the 6DoF pose of the spacecraft. Stream-2 dataset contains 100+ trajectories (each containing 300 RGB images) of the target satellite “Proba-2” generated under a realistic space simulation environment.

To make a request for the data, please use the following contact form.

Related Publication

  • Rathinam, Arunkumar, Gaudilliere, Vincent, Mohamed Ali, Mohamed Adel, Ortiz Del Castillo, Miguel, Pauly, Leo, & Aouada, Djamila. (2022b). SPARK 2022 Dataset: Spacecraft Detection and Trajectory Estimation. Zenodo. https://doi.org/10.5281/ZENODO.6599762
  • Pauly, Leo, Michele Lynn Jamrozik, Miguel Ortiz Del Castillo, Olivia Borgue, Inder Pal Singh, Mohatashem Reyaz Makhdoomi, Olga-Orsalia Christidi-Loumpasefski, et al. “Lessons from a Space Lab — An Image Acquisition Perspective.” arXiv, August 18, 2022. http://arxiv.org/abs/2208.08865.