Computer Vision in Sports

The first book of its kind devoted to this topic, this comprehensive text/reference presents state-of-the-art research and reviews current challenges in the application of computer vision to problems in sports.

Opening with a detailed introduction to the use of computer vision across the entire life-cycle of a sports event, the text then progresses to examine cutting-edge techniques for tracking the ball, obtaining the whereabouts and pose of the players, and identifying the sport being played from video footage. The work concludes by investigating a selection of systems for the automatic analysis and classification of sports play.

Topics and features: describes the latest research into ball tracking, addressing the challenges posed by the presence of occlusions and the use of only a small number of cameras; reviews various systems for player tracking and pose estimation; presents approaches for the improved generation of statistics and synthesis of virtual views; explores the “higher level” analysis of sports, from identifying types of sports to recognizing particular team behaviors based on multiple event or motion detections; discusses the detection of specific kinds of events for automatic highlights generation or searching of video archives.

The insights provided by this pioneering collection will be of great interest to researchers and practitioners involved in computer vision, sports analysis and media production.

Prof. Thomas B. Moeslund is Head of Media Technology, Aalborg, and Head of the Visual Analysis of People Lab at Aalborg University, Denmark. Dr. Graham Thomas leads the Immersive and Interactive Content team at BBC Research & Development, London, UK. Prof. Adrian Hilton is Director of the Centre for Vision, Speech and Signal Processing at the University of Surrey, Guildford, UK.

Introduction to the Use of Computer Vision in Sports
Graham Thomas, Thomas B. Moeslund and Adrian Hilton

Part I: Where is the Ball?

Ball Tracking for Tennis Video Annotation
Fei Yan, William Christmas and Josef Kittler

Plane Approximation-Based Approach for 3D Reconstruction of Ball Trajectory for Performance Analysis in Table Tennis
Sho Tamaki and Hideo Saito

On-Field Testing and Evaluation of a Goal-Line Technology System
Paolo Spagnolo, Pier Luigi Mazzeo, Marco Leo, Massimiliano Nitti, Ettore Stella and Arcangelo Distante

Part II: Where Are the Players?

Occlusion Detection via Structured Sparse Learning for Robust Object Tracking
Tianzhu Zhang, Bernard Ghanem, Changsheng Xu and Narendra Ahuja

Detecting and Tracking Sports Players with Random Forests and Context-Conditioned Motion Models
Jingchen Liu and Peter Carr

Geometry Reconstruction of Players for Novel-View Synthesis of Sports Broadcasts
Tiberiu Popa, Marcel Germann, Remo Ziegler, Richard Keiser and Markus Gross

Estimating Athlete Pose from Monocular TV Sports Footage
Mykyta Fastovets, Jean-Yves Guillemaut and Adrian Hilton

Part III: What Are They Playing?

Action Recognition in Realistic Sports Videos
Khurram Soomro and Amir Roshan Zamir

Classi?cation of Sports Types Using Thermal Imagery
Rikke Gade and Thomas B. Moeslund

Event-Based Sports Videos Classi?cation Using HMM Framework
Shyju Wilson, C. Krishna Mohan and K. Srirama Murthy

Part IV: What's Going On?

Representing Team Behaviours from Noisy Data Using Player Role
Alina Bialkowski, Patrick Lucey, Peter Carr, Sridha Sridharan and Iain Matthews

Recognizing Team Formation in American Football
Indriyati Atmosukarto, Bernard Ghanem, Mohamed Saadalla and Narendra Ahuja

Real-Time Event Detection in Field Sport Videos
Rafal Kapela, Kevin McGuinness, Aleksandra Swietlicka and Noel E. O’Connor