Image-based detection, recognition and tracking of objects are challenging tasks for any robotic system due to the enormous variety of existing objects, combined with the multitude of possibilities concerning the point of view and the object orientation. In the outside setting, illumination changes bring an extra difficulty into the equation, rendering it very difficult to reason with raw color measurements. Therefore, the Unmanned Vehicle Centre developed illumination-invariant methodologies for target detection, recognition and tracking.
An important task for unmanned robotic systems acting as search and rescue robots is the automated detection of human beings in camera images. Therefore, the Unmanned Vehicle Centre has been specifically working on human detection algorithms, mainly using visual camera data.
Gabor Marton (Budapest University of Technology and Economics)
Video Results (check also our YouTube channel):
Target tracking under changing illumination conditions
Human Victim Detection on-board of the Robudem search and rescue robot:
Incorporated in the videos