ECAI-2002 Conference Paper

[full paper]

Hypothesising Object Relations from Image Transitions

Paulo Santos, Murray Shanahan

This paper describes the construction of a qualitative spatial reasoning system based on the sensor data of a mobile robot. The robot's spatial knowledge is formalised via three sets of axioms. First of all, axioms formalising relations between pairs of spatial regions are presented. Assuming a distance function as a primitive element, the main purpose of this initial axiom set is the classification of relations between images of objects (from the robot's vision system) according to their degree of connectivity. Changes in the sensor data, due to the movement either of objects in the robot's environment or of the robot itself, are thus represented by transitions between these relations. The second set of axioms formalises these transitions. The predicates defining the transitions between image relations are related to possible interpretations for the sensor data in terms of object-observer relations; this issue is handled by the third set of axioms. These three axiom sets constitute three layers of logic-based image interpretation via abduction on transitions in the sensor data.

Keywords: Spatial Reasoning, Knowledge Representation, Cognitive Robotics

Citation: Paulo Santos, Murray Shanahan: Hypothesising Object Relations from Image Transitions. In F. van Harmelen (ed.): ECAI2002, Proceedings of the 15th European Conference on Artificial Intelligence, IOS Press, Amsterdam, 2002, pp.292-296.