Digital Sports Group

The Digital Sports Group develops pattern recognition algorithms for motion and biosignal analysis. We focus on analyzing mobile data frombody sensor networks for a wide range of applications. Research Body Sensor Networks (BSN) consist of miniature wireless sensors integrated in clothes or equipment. They acquire motion and biosignal data, and wirelessly transmit them to a central device for analysis and feedback. Motion analysis is typically performed in a lab environment. Traditional systems are large, expensive and have alimited capture volume. BSN offer the possibility to inexpensively acquire motion data in real-life situations outside of any lab. We apply pattern recognition methods to the collected data, supporting physicians and sports scientists in diagnosis and research. Biosignal analysis provides objective tools to assess the physiological state of athletes, patients and elderly people. We analyze signals like ECG, EMG or EEG to extract useful supplementary information for coaches, physicians and caregivers. Embedded algorithms for mobile applications are challenging due to limited device resources and real-time constraints. We explore and optimize pattern recognition techniques for efficient usage of memory and computational resources. This enables innovative solutions for a variety of embedded applications.

Movement classification

  • monitoring and activity recognition
  • statistics of patients and athletes

Embedded classification toolbox

  • accuracy and complexity analysis
  • target-based algorithm selection

Mobile gait analysis

  • outdoor sports classification studies
  • diagnosis of movement disorders

Virtual reality systems for athletes

  • assessment of 3D stereo vision
  • training in virtual environments

BSN consisting of body-worn sensors and central devices for analysis and feedback.