ACQUISITION AND INTERPRETATION OF 3-D SENSOR DATA FROM TOUCH

Acquisition of 3-D scene information has focused on either passive 2-D imaging methods (stereopsis, structure from motion, etc.) or 3-D range sensing methods (structured lighting, laser scanning, etc.). Little work has been done in using active touch sensing with a multifingered robotic hand to acquire scene descriptions, even though it is a well developed human capability. Touch sensing differs from other more passive sensing modalities such as vision in a number of ways. A multifingered robotic hand with touch sensors can probe, move, and change its environment. This imposes a level of control on the sensing that makes it typically more difficult than traditional passive sensors in which active control is not an issue. Second, touch sensing generates far less data than vision methods; this is especially intriguing in light of psychological evidence that shows that humans can recover shape and a number of other object attributes very reliably using touch alone. Future robotic systems will need to use dextrous robotic hands for tasks such as grasping, manipulation, assembly, inspection, and object recognition. This paper describes our use of touch sensing as part of a larger system we are building for 3-D shape recovery and object recognition using touch and vision methods. It focuses on three exploratory procedures we have built to acquire and interpret sparse 3-D touch data: grasping by containment, planar surface exploration, and surface contour exploration. Experimental results for each of these procedures are presented. © 1990 IEEE