Salt Lake City, UT, USA
Mar. 18, 2009 to Mar. 20, 2009
Raphael Hover , Computer Vision Lab - ETH Zurich, Sternwartstr. 7, CH-8092, Switzerland
Massimiliano Di Luca , Max Planck Institute for Biological Cybernetics, Spemannstr. 41, D-72076 T
Gabor Szekely , Computer Vision Lab - ETH Zurich, Sternwartstr. 7, CH-8092, Switzerland
Matthias Harders , Computer Vision Lab - ETH Zurich, Sternwartstr. 7, CH-8092, Switzerland
Data-driven haptic rendering requires processing of raw recorded signals, which leads to high computational effort for large datasets. To achieve real-time performance, one possibility is to reduce the parameter space of the employed interpolation technique, which generally decreases the accuracy in the rendering. In this paper, we propose a method for guiding this parameter reduction to maintain high accuracy with respect to the just noticeable difference for forces. To this end, we performed a user study to estimate this perception threshold. The threshold is used to assess the final error in the rendered forces as well as for the parameter reduction process. Comparison with measured data from real object interactions confirms the accuracy of our method and highlights the reduced computational effort.
Raphael Hover, Massimiliano Di Luca, Gabor Szekely, Matthias Harders, "Computationally efficient techniques for data-driven haptic rendering", WHC, 2009, World Haptics Conference, World Haptics Conference 2009, pp. 39-44, doi:10.1109/WHC.2009.4810814