Towards automatic registration of range maps

The range map registration (or alignment) phase is the main bottleneck in the 3D scanning pipeline due to the amount of user intervention required. The standard approach to registration consists of an initial rough alignment followed by an automatic refining technique (ICP). Automatic registration is an active research area, since it is the missing component to fully automatise the scanning process. Although many efforts have been made in order to avoid the manual placement, none of the existing technologies seems to be adequately accurate and flexible. This paper addresses the problem of providing a good initial guess to the ICP algorithm, by tracking the 3D-scanner motion using well known (and low cost) computer vision techniques. The proposed technique is shown to be effective and to generate sufficiently accurate guesses in the majority of the possible configurations.