Acquisition of tactile information by vision-based tactile sensor for dexterous handling of robot hands

In this paper, we propose a new method for calculation of the contact region and the location of an object from the data acquired from vision-based tactile sensor, which touch pad is an elastic transparent membrane of silicon rubber with dotted pattern printed on its inner side. The membrane is filled with translucent red colored water. The algorithm for calculation of the contact region is based on the curvature radius of the elastic membrane and considerations of the conditions for balance between the tensional forces of the membrane and the inner pressure. The location of the object is calculated from the relations between the translational displacement of the object and the displacement of the contacting dots. The rotation angles of the object are estimated from the rotation matrix of the three basis vectors by accumulating changes of the rotation angles of the rotation matrix for applying to the object rolling on the sensor. This paper also provides some experimental results that demonstrate the usefulness of the proposed method. Proposed method can be applied in many robot tasks that require dexterity of the robot hands.