A non-linear transform technique for a camera and laser 3-D scanner

3-D modeling of life-size objects is a growing market with many applications. Wright Laboratory is investigating techniques that will enhance automation of 3-D scanning systems. To assist that effort, Wright Laboratory, in cooperation with Wright State University, has built a flexible 3-D scanner platform. After mounting a camera and laser on the platform, new techniques were developed for camera-laser alignment and calibration, 3-D scanning, and 3-D surface point calculation. Of particular interest is the novel algorithm for computing 3-D surface points. Given a camera and laser pointing at a 3-D object, triangulation can calculate the 3-D surface points. The calibration procedure used by others to do this is not completely reliable in that the linear models used in camera lens distortion and geometry assumptions do not always accurately reflect what is happening in the non-linear real world. Here, we introduce a technique that compensates for the real-world application by doing a system-wide nonlinear transform. Speed is added to the triangulation technique by the use of a look-up table for range and height determination. This paper provides an introduction to 3-D scanning, a summary of current camera and laser 3-D scanning approaches, a description of the Wright Laboratory 3-D scanner, and an introduction to a novel non-linear 3-D acquisition algorithm along with its alignment and calibration procedure.