Quantitative analysis and evaluation of research on the application of computer vision in sports since the 21st century

Chen Cheng^1*Jiaxin Xue¹Wenlin Gou²Mengning Xie^3*

• ¹Harbin Sport University, Harbin, China
• ²Université de Montpellier, Montpellier, Languedoc-Roussillon, France
• ³East China University of Technology, Nanchang, Jiangxi Province, China

The purpose of this paper is to reveal the field's temporal, disciplinary, geographic, journal, and collaborative characteristics and summarize research themes and future trends to promote a systematic understanding of the field within the academic community.To identify research trends, a bibliometric analysis of 1209 publications retrieved from the Science Citation Index Expanded in the Web of Science core database was conducted.In terms of time series, publications in the field grew slowly until 2014, and publications in the field increased significantly after 2015, with polynomial models predicting 206, 233, and 263 annual publications over the next three years. In terms of disciplinary structure, three frontier disciplines utilized interdisciplinary knowledge provided by four basic disciplines to make cuttingedge breakthroughs. Geospatially, there is a three-way split between China, the United States, and the United Kingdom, while most African countries are not involved in the research. Regarding journal distribution, research in this field was published in five Q2 and four Q1 journals, mainly in computing, with fewer contributions from sports journals. The study identifies five principal research themes: skill optimization, health monitoring and injury prevention, physical performance assessment, tactical analysis and referee assistance, and immersive event experiences. Furthermore, it highlights existing research limitations and outlines directions for future exploration.Computer vision research in sports has shown high explosive growth in recent years.The field is interdisciplinary but lacks collaboration among interdisciplinary research teams. The quality of the journals published is high, but the main focus is on computer-based journals. The theme of research in this field is centered on the fundamental characteristic of serving human beings.