Building 3-D Models from Unregistered Range Images

Abstract In this paper, we describe a new approach for building a three-dimensional model from a set of range images. The approach is able to build models of free-form surfaces obtained from arbitrary viewing directions, with no initial estimate of the relative viewing directions. The approach is based on building discrete meshes representing the surfaces observed in each of the range images, mapping each of the meshes to a spherical image, and computing the transformations between the views by matching the spherical images, The meshes are built using an iterative fitting algorithm previously developed; the spherical images are built by mapping the nodes of the surface meshes to the nodes of a reference mesh on the unit sphere and by storing a measure of curvature at every node. We describe the algorithms used for building such models from range images and for matching them. We show results obtained using range images of complex objects.