Abstract | ||
---|---|---|
The paper describes a deep network based system specialized for ball detection in long shot videos. System comprises of flexible detector and classical particle tracking. The core contribution is incorporation of hy-percolumn concept in the processing pipeline achieving real-time tracking on 12MPx videos. System achieves state-of-the-art results in ISSIA-CNR Soccer Dataset and its feasibility has been tested on 4 camera prototype system. |
Year | DOI | Venue |
---|---|---|
2019 | 10.23919/MVA.2019.8757880 | 2019 16th International Conference on Machine Vision Applications (MVA) |
Keywords | Field | DocType |
BallTrack,Football ball,real-time CCTV systems,deep network,system specialized,ball detection,long shot videos,flexible detector,classical particle tracking,core contribution,processing pipeline,real-time tracking,12MPx videos,state-of-the-art results,4 camera prototype system,hypercolumn concept,ISSIA-CNR soccer dataset | Computer vision,Football,Ball tracking,Artificial intelligence,Detector,Mathematics | Conference |
ISBN | Citations | PageRank |
978-1-7281-0925-1 | 0 | 0.34 |
References | Authors | |
4 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Jacek Komorowski | 1 | 4 | 4.13 |
grzegorz kurzejamski | 2 | 3 | 3.10 |
Grzegorz Sarwas | 3 | 9 | 5.67 |