Depth Cube-Based Six Degree-of-Freedom Haptic Rendering for Rigid Bodies.

ABSTRACT

This paper presents a depth image-based six degree-of-freedom (6-DOF) haptic rendering method for rigid bodies. In order to increase the collision detection speed, a depth cube structure is developed, consisting of six-sided orthogonal depth images extracted from six virtual depth cameras located around and fixed with respect to the virtual object. With the depth cube and a set of sample points for the tool object, a quick point-in-object test can be performed. A new local occupancy map instance (LOMI) is also proposed to efficiently search for the ideal proxy location. The proposed approach works with any rigid static or dynamic virtual object, regardless of object data representation type and complexity, and achieves multi-region contacts in real-time. Several experiments subsequently confirm the effectiveness of the proposed algorithm.