Multi-Peak Estimation for Real-Time 3D Ping-Pong Ball Tracking with Double-Queue Based GPU Acceleration

Ziwei DENG; Yilin HOU; Xina CHENG; Takeshi IKENAGA

doi:10.1587/transinf.2017MVP0010

Multi-Peak Estimation for Real-Time 3D Ping-Pong Ball Tracking with Double-Queue Based GPU Acceleration

Ziwei DENG, Yilin HOU, Xina CHENG, Takeshi IKENAGA

Full Text Views

0

Cite this

Summary :

3D ball tracking is of great significance in ping-pong game analysis, which can be utilized to applications such as TV contents and tactic analysis, with some of them requiring real-time implementation. This paper proposes a CPU-GPU platform based Particle Filter for multi-view ball tracking including 4 proposals. The multi-peak estimation and the ball-like observation model are proposed in the algorithm design. The multi-peak estimation aims at obtaining a precise ball position in case the particles' likelihood distribution has multiple peaks under complex circumstances. The ball-like observation model with 4 different likelihood evaluation, utilizes the ball's unique features to evaluate the particle's similarity with the target. In the GPU implementation, the double-queue structure and the vectorized data combination are proposed. The double-queue structure aims at achieving task parallelism between some data-independent tasks. The vectorized data combination reduces the time cost in memory access by combining 3 different image data to 1 vector data. Experiments are based on ping-pong videos recorded in an official match taken by 4 cameras located in 4 corners of the court. The tracking success rate reaches 99.59% on CPU. With the GPU acceleration, the time consumption is 8.8 ms/frame, which is sped up by a factor of 98 compared with its CPU version.

Publication: IEICE TRANSACTIONS on Information Vol.E101-D No.5 pp.1251-1259

Publication Date: 2018/05/01

Publicized: 2018/02/16

Online ISSN: 1745-1361

DOI: 10.1587/transinf.2017MVP0010

Type of Manuscript: Special Section PAPER (Special Section on Machine Vision and its Applications)

Category: Machine Vision and its Applications

Authors

Ziwei DENG
  Waseda University
Yilin HOU
  Waseda University
Xina CHENG
  Waseda University
Takeshi IKENAGA
  Waseda University

Keyword

3D ball tracking, GPU acceleration, heterogeneous computing, OpenCL, particle filter, sports analysis

Cite this

Copy

Ziwei DENG, Yilin HOU, Xina CHENG, Takeshi IKENAGA, "Multi-Peak Estimation for Real-Time 3D Ping-Pong Ball Tracking with Double-Queue Based GPU Acceleration" in IEICE TRANSACTIONS on Information, vol. E101-D, no. 5, pp. 1251-1259, May 2018, doi: 10.1587/transinf.2017MVP0010.
Abstract: 3D ball tracking is of great significance in ping-pong game analysis, which can be utilized to applications such as TV contents and tactic analysis, with some of them requiring real-time implementation. This paper proposes a CPU-GPU platform based Particle Filter for multi-view ball tracking including 4 proposals. The multi-peak estimation and the ball-like observation model are proposed in the algorithm design. The multi-peak estimation aims at obtaining a precise ball position in case the particles' likelihood distribution has multiple peaks under complex circumstances. The ball-like observation model with 4 different likelihood evaluation, utilizes the ball's unique features to evaluate the particle's similarity with the target. In the GPU implementation, the double-queue structure and the vectorized data combination are proposed. The double-queue structure aims at achieving task parallelism between some data-independent tasks. The vectorized data combination reduces the time cost in memory access by combining 3 different image data to 1 vector data. Experiments are based on ping-pong videos recorded in an official match taken by 4 cameras located in 4 corners of the court. The tracking success rate reaches 99.59% on CPU. With the GPU acceleration, the time consumption is 8.8 ms/frame, which is sped up by a factor of 98 compared with its CPU version.
URL: https://global.ieice.org/en_transactions/information/10.1587/transinf.2017MVP0010/_p

Copy

@ARTICLE{e101-d_5_1251,
author={Ziwei DENG, Yilin HOU, Xina CHENG, Takeshi IKENAGA, },
journal={IEICE TRANSACTIONS on Information},
title={Multi-Peak Estimation for Real-Time 3D Ping-Pong Ball Tracking with Double-Queue Based GPU Acceleration},
year={2018},
volume={E101-D},
number={5},
pages={1251-1259},
abstract={3D ball tracking is of great significance in ping-pong game analysis, which can be utilized to applications such as TV contents and tactic analysis, with some of them requiring real-time implementation. This paper proposes a CPU-GPU platform based Particle Filter for multi-view ball tracking including 4 proposals. The multi-peak estimation and the ball-like observation model are proposed in the algorithm design. The multi-peak estimation aims at obtaining a precise ball position in case the particles' likelihood distribution has multiple peaks under complex circumstances. The ball-like observation model with 4 different likelihood evaluation, utilizes the ball's unique features to evaluate the particle's similarity with the target. In the GPU implementation, the double-queue structure and the vectorized data combination are proposed. The double-queue structure aims at achieving task parallelism between some data-independent tasks. The vectorized data combination reduces the time cost in memory access by combining 3 different image data to 1 vector data. Experiments are based on ping-pong videos recorded in an official match taken by 4 cameras located in 4 corners of the court. The tracking success rate reaches 99.59% on CPU. With the GPU acceleration, the time consumption is 8.8 ms/frame, which is sped up by a factor of 98 compared with its CPU version.},
keywords={},
doi={10.1587/transinf.2017MVP0010},
ISSN={1745-1361},
month={May},}

Copy

TY - JOUR
TI - Multi-Peak Estimation for Real-Time 3D Ping-Pong Ball Tracking with Double-Queue Based GPU Acceleration
T2 - IEICE TRANSACTIONS on Information
SP - 1251
EP - 1259
AU - Ziwei DENG
AU - Yilin HOU
AU - Xina CHENG
AU - Takeshi IKENAGA
PY - 2018
DO - 10.1587/transinf.2017MVP0010
JO - IEICE TRANSACTIONS on Information
SN - 1745-1361
VL - E101-D
IS - 5
JA - IEICE TRANSACTIONS on Information
Y1 - May 2018
AB - 3D ball tracking is of great significance in ping-pong game analysis, which can be utilized to applications such as TV contents and tactic analysis, with some of them requiring real-time implementation. This paper proposes a CPU-GPU platform based Particle Filter for multi-view ball tracking including 4 proposals. The multi-peak estimation and the ball-like observation model are proposed in the algorithm design. The multi-peak estimation aims at obtaining a precise ball position in case the particles' likelihood distribution has multiple peaks under complex circumstances. The ball-like observation model with 4 different likelihood evaluation, utilizes the ball's unique features to evaluate the particle's similarity with the target. In the GPU implementation, the double-queue structure and the vectorized data combination are proposed. The double-queue structure aims at achieving task parallelism between some data-independent tasks. The vectorized data combination reduces the time cost in memory access by combining 3 different image data to 1 vector data. Experiments are based on ping-pong videos recorded in an official match taken by 4 cameras located in 4 corners of the court. The tracking success rate reaches 99.59% on CPU. With the GPU acceleration, the time consumption is 8.8 ms/frame, which is sped up by a factor of 98 compared with its CPU version.
ER -