Analyzing volleyball videos based on 3D ball trajectories was relatively overlooked before. In this paper, we focus on how ball trajectories can benefit volleyball video analysis. Based on videos captured by two cameras from different viewpoints, we detect the volleyball and construct 3D ball trajectories. We then propose a trajectory segmentation and classification method based on BERT (Bidirectional Encoder Representation for Transformer). The volleyball at each frame can be categorized into one of six trajectory classes, e.g., serve and attack, and a long ball trajectory showing the ball being hit back and forth is appropriately segmented. We believe that this is a very first study adopting the language model technique to analyze ball trajectories, and results of trajectory segmentation and classification can enable more advanced volleyball analysis.
© Copyright 2021 IEEE Transactions on Circuits and Systems II: Express Briefs. All rights reserved.