Robotic perception group (ROP) deals with machine perception, namely computer vision, tactile perception for robotic manipulation, perception from other sensory modalities, and cognitive approach to human-robot interaction. In computer vision we deal with object segmentation using graph methods, videosequences analysis, object detection in images and videos. The other area of interest is the vision-based quality check for industry.
In the automotive domain, we perceive the traffic from the outside of a car and also from the ego-car (project UP-Drive). In robotic applications, we are interested in the situation awareness around a mobile robot (project TRADR). In dual-arm manipulation with soft objects as pieces of a garment (project CloPeMa finished in 2015, followed by the project RadioRoSo) and perception in the Industry 4.0 context.