Haptische Objekterkennung mit einer humanoiden Roboterhand

Abstract

The focus of this work is on the analysis of haptic sensor data acquired by a humanoid robot for shape-based object classification. The central theme is the examination of different approaches for the fusion of haptic sensor data in order to distinguish objects roughly with a few or in detail with several samples. A part of the presented system is an attention space with provide a strategy for an active object classification.