Tracking in real-time rigid, articulated and deformable targets such as people, body parts and objects, is a very important task for the perception-cognition-action loop of technical systems.
Nowadays tracking different rigid, articulated and deformable targets (such as people, body parts and objects) in real-time is a very important task for the perception-cognition-action loop of technical systems. Example application scenarios include, but are not limited to, side by side human-machine-interaction, autonomous (re-)acting of robots based on their environment, and enforcing safety requirements.
At the chair for Robotics and Embedded Systems, a modular and general-purpose tracking framework has been developed: the OpenTL library. OpenTL allows rapid development of any model-based tracking application, using a wide range of state-of-the-art algorithms and computer vision methodologies. Some features include: multi-target detection, Bayesian filters, multi-camera tracking, multi-modal sensor fusion, seamless GPU-accelerated code, and many others as well.
For more information, please visit the related CoTeSys page (ITrackU Project): www6.in.tum.de/Main/ResearchItracku