Authors: Alexander Hiemann , Thomas Kautz , Johannes Waschke , Mario Hlawitschka
Abstract: Sensor-based data acquisition is an important aspect of training methodology in many different types of sports. Especially the analysis of data gathered from sensor-based positioning systems or motion sensors is a relevant source of information that is used to achieve performance optimization and to generate competitive advantages. Many recent approaches address the subject of sensor-fusion of athletes position data and high frequent motion data gathered from inertial measurement units (IMU) in order to enhance position accuracy and to increase the temporal resolution of position data. Besides this “technological sensor fusion” approach, data from both data sources can be integrated into a combined multi-modal visualization in order to provide deeper insights for training theory and to allow for further in-depth analysis. In this paper, we introduce a visualization-based information fusion approach that combines the insights from both data sources without the need for extensive technological integration. In addition, this visualization technique improves the interpretability of inertial data for athletes and coaches by establishing the relation to the spatial context of the position data. We use this visualization technique for a detailed analysis of motion data of athletes and accordingly to derive a detailed step analysis of skaters in short track and speed skating.