3-D Objekterkennung und Szeneninterpretation: Ein System zur multimodalen Beschreibung von Innenraumszenen

Abstract

Cognitive Vision as a field of Computer Vision among other tasks deals with the integration of the perceptive component in a holistic cognitive system. The cognitive vision system ORCC presented here uses cognitively motivated methods for recognition that are partially independent of the kind of the sensor. The system provides a functional and textual scene description that also takes alternative interpretations into account which are used for the subsequent cognitive processing within a speech-interaction module. Thus the difficult but indispensable process of man-machine interaction that itself is a prerequisite for the man-machine learning process is represented within an indoor scene. For this task methods for spatial calculi from the field of Spatial Cognition in an ontology based realization are used. ORCC combines diverse recognition strategies that afford an extensive description of an unreserved scene: In a first step the room demarcations and structurally simple objects such as tables are extracted using as well functional as structural properties. Then further objects are segmented based on their position, followed by a structurally more complex and a more shape-oriented recognition step. Then, this spatial information is enriched with colour-based information about the objects. Afterwards, the resulting scene description can be used as an input for a speech-based man-machine dialogue about the objects within in the scene. The advantage of the ORCC-System is the combination of a 2-D colour intensity camera and a 3-D laser range scanner. Thus more powerful 3-D data can be used for central segmentation and classification tasks and be enriched with 2-D properties like e.g. colour.