US 11,861,806 B2
End-to-end camera calibration for broadcast video
Long Sha, Brisbane (AU); Sujoy Ganguly, Chicago, IL (US); and Patrick Joseph Lucey, Chicago, IL (US)
Assigned to STATS LLC, Chicago, IL (US)
Filed by STATS LLC, Chicago, IL (US)
Filed on Apr. 9, 2021, as Appl. No. 17/226,205.
Claims priority of provisional application 63/008,184, filed on Apr. 10, 2020.
Prior Publication US 2021/0319587 A1, Oct. 14, 2021
Int. Cl. G06T 3/00 (2006.01); G06T 7/80 (2017.01); G06T 7/11 (2017.01); G06V 10/82 (2022.01); H04N 21/854 (2011.01); G06N 3/08 (2023.01); G06V 20/40 (2022.01); G06V 30/262 (2022.01); G06F 18/214 (2023.01); G06V 30/19 (2022.01); G06V 10/24 (2022.01)
CPC G06T 3/00 (2013.01) [G06F 18/214 (2023.01); G06N 3/08 (2013.01); G06T 3/0093 (2013.01); G06T 7/11 (2017.01); G06T 7/80 (2017.01); G06V 10/24 (2022.01); G06V 10/82 (2022.01); G06V 20/49 (2022.01); G06V 30/19173 (2022.01); G06V 30/274 (2022.01); H04N 21/854 (2013.01); G06T 2207/10016 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30244 (2013.01)] 20 Claims
OG exemplary drawing
 
8. A system for calibrating a broadcast video feed, comprising:
a processor; and
a memory having programming instructions stored thereon, which, when executed by the processor, performs one or more operations, comprising:
receiving a target broadcast video feed for a target sporting event;
partitioning the target broadcast video feed into a plurality of target frames;
for each target frame in the plurality of target frames, generating, by a neural network, a target homography matrix, wherein the neural network is pre-trained, the generating comprising:
generating a playing surface agnostic view of a playing surface in each target frame,
identifying a template that most closely matches the playing surface agnostic view of the playing surface,
retrieving a template homography matrix associated with the template,
predicting a relative homography matrix based on the template and the playing surface agnostic view, and
generating the target homography matrix based on the template homography matrix and the relative homography matrix; and
calibrating the target broadcast video feed by warping each target frame by a respective target homography matrix.