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

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 multiview 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.

View Priority Based Threads Allocation and Binary Search Oriented Reweight for GPU Accelerated Real-Time 3D Ball Tracking

“…As to the chosen algorithm, because of the prohibitively large number of simulations which is also known as particles, the extreme large computation complexity is shown as a problem in achieving real-time performances. An extension work [24] of the chosen algorithm [9] on ping-pong ball tracking and its accelerated work [28] have shown that algorithms basis on particle filter have large potential for acceleration. Basing on the characteristic of ping-pong ball during games, work [24] makes some improvements in the observation and the resampling algorithms of the chosen one [9] which are not friendly to other ball games.…”

“…Basing on the characteristic of ping-pong ball during games, work [24] makes some improvements in the observation and the resampling algorithms of the chosen one [9] which are not friendly to other ball games. What's more, the modification has caused a considerable distinction in accelerated structure [28] on GPU system which means that it [28] cannot be applied in the acceleration of the chosen algorithm [9]. The goal of this research is implementing a 3D ball tracking system to obtain both high accuracy and real-time realization based on the chosen algorithm [9].…”