Determining Both 3D Position and Orientation from Projection of Rhombic Pseudo-Random Pattern

Z. Song and R. Chung (PRC)

Keywords

Structured light based system, 3D reconstruction, pseudo random pattern

Abstract

3D position and orientation are two principal descriptions of object shape. Yet orientation information, if merely constructed from position information, will not provide information beyond what is already contained in the position data. In this article, we describe a design of structured light system that determines both position and orientation independently. The system involves projection of a specially designed pseudo-random pattern. In traditional designs the centroids of the pattern elements are used as the feature points. We use the grid points between pattern elements instead as the feature points. The grid-points have the essence that, not only do they allow 3D positions to be determined at the points, they also allow 3D orientations to be extracted from the grid lines that form the points. A mechanism of determining surface orientations from the imaged grid-lines is described. All the above can be achieved with only a single pattern illumination and image capture. Extensive experiments are presented to show the performance of the design.

Important Links:



Go Back