Player trajectory generation via multiple camera player tracking

Patent metadata

Field	Value
Publication number	US-11967086-B2
Application number	US-201917435923-A
Country	US
Kind code	B2
Filing date	Jul 31, 2019
Priority date	Jul 31, 2019
Publication date	Apr 23, 2024
Grant date	Apr 23, 2024

Abstract

A method for trajectory generation based on player tracking is described herein. The method includes determining a temporal association for a first player in a captured field of view and determining a spatial association for the first player. The method also includes deriving a global player identification based on the temporal association and the spatial association and generating a trajectory based on the global player identification.

First claim

What is claimed is:

1. A method for trajectory generation based on player tracking, comprising:
determining a spatial association for a player in a captured field of view;
determining a temporal association for the player in the captured field of view;
deriving a global player identification (ID) based on the temporal association and the spatial association, the global player ID associated with a jersey number of the player; and
generating a trajectory based on the global player ID.

2. The method of claim 1 , wherein the temporal association is found by determining a bounding box including the player in multiple frames of a same camera view of the captured field of view.

3. The method of claim 1 , wherein the spatial association is found by determining a player location across a plurality of camera views at a particular timestamp.

4. The method of claim 1 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of short tracklets for the player.

5. The method of claim 1 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of the short tracklets using the jersey number of the player and a team tag of the player and eliminating outliers by a motion constraint.

6. The method of claim 1 , further including applying global smoothing to the generated trajectory.

7. The method of claim 1 , further including constructing a virtual camera within a three-dimensional volumetric representation of the captured field of view, and progressing through the three-dimensional volumetric representation according to the generated trajectory.

8. The method of claim 1 , further including deriving the spatial association at a frame-level to associate a plurality of players from each camera view and computing a three-dimensional position for each player using projective matrices.

9. The method of claim 1 , further including assigning the global player ID the jersey number and a team tag.

10. The method of claim 1 , further including, in response to an indication of partial or total occlusion of the player according to a distance matrix, allocating a new global player ID for the player.

11. A system for trajectory generation based on player tracking, comprising:
a single camera player tracking (SCT) module to determine a temporal association for a first player in a captured field of view;
a multiple camera association (MCA) module to determine a spatial association for the first player;
a multiple camera tracking (MCT) module to derive a global player identification (ID) based on the temporal association and the spatial association, the global player ID associated with a jersey number of the first player; and
a player trajectory optimization (PTO) module to generate a trajectory based on the global player ID.

12. The system of claim 11 , wherein the temporal association is found by determining a bounding box including the first player in multiple frames of a same camera view of the captured field of view.

13. The system of claim 11 , wherein the spatial association is found by determining a player location across a plurality of camera views at a particular timestamp.

14. The system of claim 11 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of the short tracklets for the first player.

15. The system of claim 11 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of the short tracklets using the jersey number of the first player and a team tag of the first player and eliminating outliers by a motion constraint.

16. The system of claim 11 , wherein the PTO module is to apply global smoothing to the generated trajectory.

17. The system of claim 11 , further including an immersive viewing manager to construct a virtual camera within a three-dimensional volumetric representation of the captured field of view, and progress through the three-dimensional volumetric representation according to the generated trajectory.

18. The system of claim 11 , wherein the MCA module is to derive the spatial association at a frame-level to associate a plurality of players from each camera view and compute a three-dimensional position for each player using projective matrices.

19. The system of claim 11 , wherein the MCT module is to assign the global player ID the jersey number and a team tag.

20. The system of claim 11 , wherein in response to an indication of partial or total occlusion of the first player according to a distance matrix, the MCT module is to allocate a new global player ID for the first player.

21. At least one non-transitory computer-readable medium, comprising instructions to direct a processor to:
determine a spatial association for a player a captured field of view;
determine a temporal association for the player in the captured field of view;
derive a global player identification (ID) based on the temporal association and the spatial association, the global player ID associated with a jersey number of the player; and
generate a trajectory based on the global player ID.

22. The non-transitory computer readable medium of claim 21 , wherein the temporal association is found by determining a bounding box including the player in multiple frames of a same camera view of the captured field of view.

23. The non-transitory computer readable medium of claim 21 , wherein the spatial association is found by determining a player location across a plurality of camera views at a particular timestamp.

24. The non-transitory computer readable medium of claim 21 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of the short tracklets for the player.

25. The non-transitory computer readable medium of claim 21 , wherein short tracklets are generated based on the global player ID, and the trajectory is generated by concatenating a plurality of the short tracklets using the jersey number of the player and a team tag of the player and outliers are eliminated by a motion constraint.

Assignees

• Intel Corp

Inventors

• Fang Yikai
• Li Qiang
• Li Wenlong
• Liu Chenning
• Ling Chen
• Tao Hongzhi
• Wang Yumeng
• Zheng Hang

Classifications

• G06T7/246 Primary

  using feature-based methods, e.g. the tracking of corners or segments · CPC title

• A63F13/213

  comprising photodetecting means, e.g. cameras, photodiodes or infrared cells (A63F13/219 takes precedence) · CPC title

• A63F13/42

  by mapping the input signals into game commands, e.g. mapping the displacement of a stylus on a touch screen to the steering angle of a virtual vehicle · CPC title

• A63F13/65

  automatically by game devices or servers from real world data, e.g. measurement in live racing competition · CPC title

• G06T7/292

  Multi-camera tracking · CPC title