An image interpretation software architecture for self-orientation of robots

Presents a software architecture for a computer vision approach to help a mobile robot to determine its own position and orientation in manufacturing environments without special optical landmarks. Furthermore the identification and detection of unexpected obstacles is supported. The approach follows the blackboard paradigm in order to achieve a higher modularity and extendability. Knowledge sources for modelling, image processing and controlling are integrated in the system and perform the necessary operations to determine the orientation.<>