A new method for the registration of three-dimensional point-sets: The Gaussian Fields framework

Faysal Boughorbel, Muharrem Mercimek, Andreas Koschan, Mongi Abidi
2010 Image and Vision Computing  
In this paper, we present a 3D automatic registration method based on Gaussian Fields and energy minimization. A continuously differentiable energy function is defined, which is convex in a large neighborhood of the alignment parameters. We show that the size of the region of convergence can be significantly extended reducing the need for close initialization and overcoming local convergence problems of the standard Iterative Closest Point (ICP) algorithms. Moreover, the Gaussian criterion can
more » ... e applied with linear computational complexity using Fast Gauss Transform methods. Experimental evaluation of the technique using synthetic and real datasets demonstrates the usefulness as well as the limits of the approach. Published by Elsevier B.V. 0262-8856/$ -see front matter Published by Elsevier B.V.
doi:10.1016/j.imavis.2009.05.003 fatcat:n3oxkcqmujcorl2b7ukeb4p4wq