Use of deep learning in soccer videos analysis: survey

Akan S., Varlı S.

Multimedia Systems, cilt.29, sa.3, ss.897-915, 2023 (SCI-Expanded) identifier identifier

  • Yayın Türü: Makale / Tam Makale
  • Cilt numarası: 29 Sayı: 3
  • Basım Tarihi: 2023
  • Doi Numarası: 10.1007/s00530-022-01027-0
  • Dergi Adı: Multimedia Systems
  • Derginin Tarandığı İndeksler: Science Citation Index Expanded (SCI-EXPANDED), Scopus, Academic Search Premier, PASCAL, Applied Science & Technology Source, Compendex, Computer & Applied Sciences, INSPEC
  • Sayfa Sayıları: ss.897-915
  • Anahtar Kelimeler: Soccer, Football, Survey, Review, Overview, Deep learning, Event detection, Player detection, Ball detection, Player tracking, Ball tracking, Game analysis, Team performance
  • Yıldız Teknik Üniversitesi Adresli: Evet

Özet

© 2022, The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature.The demand for video analysis has been rapidly increasing in the last decade. Video analysis plays a critical role in various technologies, including medical diagnosis, security surveillance, robotics, and sport. Soccer is the most popular sport in our culture, with millions of fans. Many video analysis approaches have been developed in recent years to assist and provide important information to spectators, referees, coaches, and players. Most of these approaches are aimed towards detecting and tracking players or the ball, event detection, and analysis of the game. For this purpose, various classical or deep learning-based strategies have been used. This study investigates deep learning-based techniques that have been proposed over the last few years to analyze football videos. The purpose of this study is not to compare current methodologies, but to show the most recent research in the field. This paper investigates the challenges of soccer video analysis and its application groups, e.g., player/ball detection and tracking, event detection, and game analysis. This paper also reviews the used deep learning-based methods, their performance, advantages, and disadvantages in soccer videos, and finally, concludes with future potential in the analysis of soccer videos.