Reverse engineering with a structured light system

This paper proposes a scanning procedure using an structured light system (SLS). To respond to the increasing needs of reverse engineering, it becomes very important to achieve high efficiency of measuring procedure. As an SLS can measure only the visible area from a specific direction, it is necessary to scan multiple times to obtain a complete model. However, it is more desirable to minimize the number of scanning directions to save the measuring time and the amount of required memory. The proposed procedure identifies missing areas from multiple range images (scanning data), and attempts to locate additional scanning orientations to fill the missing areas. The core part of the proposed procedure is the identification of missing areas from multiple range images. Since the identification of missing areas may be called repeatedly during the scanning procedure, it is very important to improve the efficiency. To satisfy the requirement, this paper proposes a new method for identifying missing areas using a graphic hardware. Once missing areas are identified, the proposed procedure computes additional scanning orientations by using the visibility information of missing areas. As the proposed procedure is based on well-known 2D geometric algorithms, it is very efficient.