Shape-Based Spatial Representation for Mobile Robots

Goal: Describe an object-centered, abstract spatial represention to be used as geometric foundation in mobile robot applications.
Characteristics of spatial representations currently used: Occupancy grids, bitmap like representations are widely used Not suitable for communication Requires additional sensor information in mapping applications (dead reckogning) Requires much data to be processed Object maps based on simple geometric features (lines, special landmarks) Object-centered representation are judged necessary to cope with changing environments Presence of features limits applicability Feature extraction and recognition lacks of reliability Geometric information remains largely uninterpreted.
Features of Shape: Shows of advantages from feature-based representations while avoiding its shortcomings: Provides an object-centered, abstract, and compact representation Representation grows by the environment's complexity, not by its size Object-centered representations are viewed a necessatity to tackle dynamic environments Is not limited to special working environments as shape is universal Algorithms can easily benefit from a more abstract representation resulting in better performance: less data needs to be processed and the data's structural organization can be exploited Bridges from metric to abstract information: Grants access to metric information, e.g., necessary for robot motion Offers a qualitative or topological view on configurations (of objects), enabling an interface to higher-level spatial reasoning Applicable in robot mapping, localization, and navigation applications
Key Techniques: Extracting shape representation from perception Shape-based tracking of objects Updating the map & localizing within (SLAM)
References: L.J. Latecki, R. Lakämper, and D. Wolter: Shape Similarity and Visual Parts. Proc. Int. Conf. on Discrete Geometry for Computer Imagery (DGCI), November 2003 L.J. Latecki and R. Lakämper: Shape Similarity Measure Based on Correspondence of Visual Parts. IEEE Trans. Pattern Analysis and Machine Intelligence (PAMI) 22(10), pp. 1185-1190, October 2000