This dataset is a large-scale satellite video moving target detection and tracking dataset. It consists of high-resolution videos taken by the Jilin-1 satellite platform. It has a resolution of 12000×5000 pixels, covers an area of several square kilometers, and includes four common types of moving targets such as airplanes, cars, ships and trains. These targets have been manually annotated with a total of 853,911 instances annotated by axis-aligned bounding boxes. The VISO dataset aims to advance the state of the art in satellite video analysis by addressing challenges such as small object size, low spatial resolution, and limited appearance and texture information. The dataset provides subsets for training, validation, and testing, where the training and validation sets contain annotation information, while the test set contains only images.

Citation

If you find our work useful in your research, please consider citing: @article{yin2021detecting, title={Detecting and Tracking Small and Dense Moving Objects in Satellite Videos: A Benchmark}, author={Yin, Qian and Hu, Qingyong and Liu, Hao and Zhang, Feng and Wang, Yingqian and Lin, Zaiping and An, Wei and Guo, Yulan} journal={IEEE Transactions on Geoscience and Remote Sensing}, year={2021}, publisher={IEEE} } @cs.ox.ac.uk if you have any questions.