Nonlocal treatment of systematic errors in the processing of sparse and incomplete sensor data
Martin Beckerman, E.M. Oblow · OSTI OAI (U.S. Department of Energy Office of Scientific and Technical Information) · 1988
A methodology has been developed for the treatment of systematic errors which arise in the processing of sparse and incomplete sensor data. We present a detailed application of this methodology to the construction of navigation maps from wide-angle sonar sensor data acquired by the HERMIES IIB mobile robot. Our uncertainty approach is explcitly nonlocal. We use a binary labelling scheme and a simple logic for the rule of combination. We then correct erroneous interpretations of the data by analyzing pixel patterns of conflict and by imposing consistent labelling conditions. 9 refs., 6 figs.