Department of Image Processing

Restoration of motion-blurred objects is a topic much less frequently addressed in the literature compared to general image or video deblurring. We are interested in a particular phenomenon, which we call fast moving objects (FMO) – objects that travel large distance during single frame exposure time and are therefore heavily blurred. We tackle two connected problems:

• Recovering the FMO appearance, shape and trajectory from a single image - Deblatting (deblurring and matting)
• Tracking FMOs in videos - Tracking by Deblatting

Publications:

• Rozumnyi D., Kotera Jan, Šroubek Filip, Matas J.: Sub-Frame Appearance and 6D Pose Estimation of Fast Moving Objects, CVPR, [2020]
• Kotera Jan: Image Deblurring in Demanding Conditions, doctoral thesis, [2020] Download
• Kotera Jan, Rozumnyi D., Šroubek Filip, Matas J.: Intra-Frame Object Tracking by Deblatting, Proceedings of the IEEE International Conference on Computer Vision 2019 (ICCV 2019), p. 2300-2309, ICCV 2019. International conference on computer vision, (Seoul, KR, 20191027) [2019] Download DOI: 10.1109/ICCVW.2019.00283
• Rozumnyi D., Kotera Jan, Šroubek Filip, Matas J.: Non-causal Tracking by Deblatting, Pattern Recognition : 41st DAGM German Conference, DAGM GCPR 2019, p. 122-135, German Conference on Pattern Recognition 2019 /44./,
  (Dortmund, DE, 20190910) [2019] Download DOI: 10.1007/978-3-030-33676-9_9
• Kotera Jan, Šroubek Filip: Motion Estimation and Deblurring of Fast Moving Objects, Proceedings of the 25th IEEE International Conference on Image Processing 2018, p. 2860-2864, ICIP 2018, [2018]
• Rozumnyi D., Kotera Jan, Šroubek Filip, Novotný L., Matas J.: The World of Fast Moving Objects, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), p. 5203-5211, The IEEE Conference on Computer Vision and Pattern Recognition (CVPR), (Honolulu, US, 20170721) [2017] Download DOI: 10.1109/CVPR.2017.514