TY - GEN
T1 - Real-time 3D ball tracking with CPU-GPU acceleration using particle filter with multi-command queues and stepped parallelism iteration
AU - Hou, Yilin
AU - Cheng, Xina
AU - Ikenaga, Takeshi
N1 - Funding Information:
ACKNOWLEDGMENT This work was supported by KAKENHI (26280016) and (16K13006).
PY - 2017/12/15
Y1 - 2017/12/15
N2 - 3D ball tracking is a critical function in manyapplications such as game and players behavior analysis, andreal time implementation has become increasingly importantfor it can be used for live broadcast and TV contents. To reacha high accuracy, algorithms usually are time consuming due toa large set of calculations which is challenging to meet realtime demanding. This paper proposes multiple commandqueues, tactical threads allocation and stepped iterativeaddition to empower such a capacity on the CPU-GPUplatform. Multiple command queues achieves a parallelismbetween tasks in the algorithm. Secondly, the tactical threadsallocation helps mapping the algorithm into GPU andenhances synchronism between threads. And this paperproposes stepped iterative addition to achieve partialparallelism in a sequential operation. This work implements inan Intel Core i7-6700 GPU and AMD Radeon R9 FURY GPU.Tracking speed of our work increases 37.8 times from original431ms to 11.7ms while the success rate of the algorithm retainsover 99%. This result fully meets the requirement of 16.6msper frame for 60fps video real-time tracking.
AB - 3D ball tracking is a critical function in manyapplications such as game and players behavior analysis, andreal time implementation has become increasingly importantfor it can be used for live broadcast and TV contents. To reacha high accuracy, algorithms usually are time consuming due toa large set of calculations which is challenging to meet realtime demanding. This paper proposes multiple commandqueues, tactical threads allocation and stepped iterativeaddition to empower such a capacity on the CPU-GPUplatform. Multiple command queues achieves a parallelismbetween tasks in the algorithm. Secondly, the tactical threadsallocation helps mapping the algorithm into GPU andenhances synchronism between threads. And this paperproposes stepped iterative addition to achieve partialparallelism in a sequential operation. This work implements inan Intel Core i7-6700 GPU and AMD Radeon R9 FURY GPU.Tracking speed of our work increases 37.8 times from original431ms to 11.7ms while the success rate of the algorithm retainsover 99%. This result fully meets the requirement of 16.6msper frame for 60fps video real-time tracking.
KW - GPU
KW - Hardware acceleration
KW - Particle filter
KW - Sports analysis
UR - http://www.scopus.com/inward/record.url?scp=85033442376&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85033442376&partnerID=8YFLogxK
U2 - 10.1109/ICMIP.2017.59
DO - 10.1109/ICMIP.2017.59
M3 - Conference contribution
AN - SCOPUS:85033442376
T3 - Proceedings - 2017 2nd International Conference on Multimedia and Image Processing, ICMIP 2017
SP - 235
EP - 239
BT - Proceedings - 2017 2nd International Conference on Multimedia and Image Processing, ICMIP 2017
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2nd International Conference on Multimedia and Image Processing, ICMIP 2017
Y2 - 17 March 2017 through 19 March 2017
ER -