| Surface Patch Reconstruction from Minimal Tactile Data (2008) | |||||||||||||||
Abstract | |||||||||||||||
| Abstract — This paper studies the reconstruction of unknown curved surfaces through finger tracking. A patch can be generated from tactile data points along three concurrent surface curves in two steps. First, the Darboux frame at the curve intersection is estimated to reflect the local geometry. Next, polynomial fitting is carried out in the frame. Minimization of the total (absolute) Gaussian curvature of the surface fitting function effectively prevents unnecessary folds otherwise expected to result from the use of such “1D ” data. The implementation involves a 2axis joystick sensor, a 3-fingered 4-DOF BarrettHand, and a 4-DOF Adept SCARA robot. Experiments have demonstrated remarkable accuracy of reconstruction. With an objective to investigate the amount of tactile information needed for shape recovery, this work has potential applications in grasping, dexterous manipulation, terrain exploration, and other close contact situations where occlusions often pose a problem for cameras and range sensors. Index Terms — Shape reconstruction, touch sensing, surface fitting, contour tracking, total Gaussian curvature. | |||||||||||||||
Publication details | |||||||||||||||
| |||||||||||||||