Wide-Ranging applications for Motion Analysis
Extracting information from motion patterns
The basis of automatic motion analysis is computer-aided object monitoring, which uses observation data (e.g., from a camera network) to reconstruct the path of a motion and extracts the information required for analysis, such as position, velocity and orientation. The motion data of individual objects can be combined to motion patterns, which are compared with motion patters that have been identified as typical in order to detect critical system states automatically.
SCCH is involved in automated motion analysis and modeling based on methods of computational intelli¬gence. This requires software technologies for detection and monitoring especially of objects that are non-rigid (e.g., persons) and deformable (e.g., organs), of ir¬regular and highly dynamic processes (e.g., of cancer cells, footballs), in (quasi) real time and under outdoor conditions.
One example of motion analysis in sports is a research project with ABATEC Electronic AG to further develop the Local Position Measurement (LPM) system. In foot¬ball and other sports, this system enables the analysis of game strategies, player positions, paths, speeds and acceleration curves in combination with biometric player data such as heart and breathing rates. “We have developed a framework that lets us filter out position data for objects or persons from video sequences. The framework builds on highly developed image process¬ing methods that enable high precision and flexibility in object monitoring and that have been optimized with respect to real-time conditions,” reports project leader Dr. Thomas Hoch. DI Günther Stelzhammer of ABATEC Electronic AG adds: “The video-based system is meanwhile very well integrated in our existing LPM system and for the first time permits more in-depth analyses in sports such as football, tennis, volleyball and floorball (indoor hockey).”