Computer supported automated sports game analysis

The notational analysis of sports games is an important tool for the research of their structure. It is the base for the science of training and training practise of the investigated type of sport. During the last decades the use of computers, digital video systems and databases improved the notational analysis substantially. New technologies like picture recognition, tracking of objects and audio analysis enable to reduce the manual effort and subjectivity during structure research. An interdisciplinary cooperation between the Institute of Sports Science at the Karl- Franzens-Universität Graz, the Institute for Computer Graphics and Vision at the Technische Universität Graz and the Institute for Information Systems and Information Management at Joanneum Research Graz will use these new technologies for the research on the structure of sports games. To fulfil the aims of the project, an existing analysis system, which has already been used for scientific research, will be augmented with the new tools. As beach volleyball has a not too complex structure (only 4 players and a little playing field) it will be an excellent test bed for the planned investigations. Already in previous projects we have split digital competition videos into before defined sequences for analysis. The video sequences have been saved in a database and used for statistics. This enabled an immediate statistical analysis as well as quick access to the related sequences. During the project the novel video and audio analysis methods shall improve the process of analysis substantially. Beside the categorization into game sequences also the movements of the players and the ball will be determined with the help of object tracking methods. The categorization according to the used sport techniques will give a comprehensive description of beach volleyball. With this information we expect to determine the most important elements which are responsible for success for failure. Beside the quantitative description, the recording of the game sequences in a database allows focused feedback for the athlete or an opponent`s analysis. The tracking of players makes it possible to analyze the physical load during a competition and will demonstrate the physical demands on the athletes. The reduction of the manual work due to the use of new technologies should enable for the first time investigating a statistical relevant amount of data.

The notational analysis of sports games is an important tool for the research of their structure. It is the base for the science of training and training practise of the investigated type of sport. During the last decades the use of computers, digital video systems and databases improved the notational analysis substantially. New technologies like picture recognition, tracking of objects and audio analysis enable to reduce the manual effort and subjectivity during structure research. An interdisciplinary cooperation between the Institute of Sports Science at the Karl- Franzens-Universität Graz, the Institute for Computer Graphics and Vision at the Technische Universität Graz and the Institute for Information Systems and Information Management at Joanneum Research Graz will use these new technologies for the research on the structure of sports games. To fulfil the aims of the project, an existing analysis system, which has already been used for scientific research, will be augmented with the new tools. As beach volleyball has a not too complex structure (only 4 players and a little playing field) it will be an excellent test bed for the planned investigations. Already in previous projects we have split digital competition videos into before defined sequences for analysis. The video sequences have been saved in a database and used for statistics. This enabled an immediate statistical analysis as well as quick access to the related sequences. During the project the novel video and audio analysis methods shall improve the process of analysis substantially. Beside the categorization into game sequences also the movements of the players and the ball will be determined with the help of object tracking methods. The categorization according to the used sport techniques will give a comprehensive description of beach volleyball. With this information we expect to determine the most important elements which are responsible for success for failure. Beside the quantitative description, the recording of the game sequences in a database allows focused feedback for the athlete or an opponent`s analysis. The tracking of players makes it possible to analyze the physical load during a competition and will demonstrate the physical demands on the athletes. The reduction of the manual work due to the use of new technologies should enable for the first time investigating a statistical relevant amount of data.