LaserGun: A tool for hybrid 3D reconstruction

Abstract

We present a tool for the acquisition of 3D textured models of objects of desktop size using an hybrid computer vision framework. This framework combines active laser-based triangulation with passive motion estimation. The 3D models are obtained by motion-based alignment (with respect to a fixed world frame) of imaged laser profiles backprojected onto time-varying camera frames. Two distinct techniques for estimating camera displacements are described and evaluated. The first is based on a Simultaneous Localization and Mapping (SLAM) approach, while the second exploits a planar pattern in the scene and recovers motion by homography decomposition. Results obtained with a custom laser-camera stereo setup - implemented with off-the-shelf hardware - show that a trade-off exists between the greater operational flexibility of SLAM and the higher model accuracy of the homography-based approach. © 2013 Springer-Verlag.