Towards an Automated True Color Projection onto Adaptively Reduced Point Data from 3D Surface Scans

3D surface scanners can only record discrete data sets (point clouds). The meshing process is a complex issue and in the last years there were lots of algorithms developed to solve this problem. In this work an algorithm will be presented, which generates a textured regular surface model from arbitrary scattered 3D scan data. A color-coded normal field, delivered with the input data, allows for a meaningful projection of curved surfaces and, therefore, for iteratively building up a highly detailed regular mesh. As most scanners are able to pointwise record true color information, such a mesh can serve as a template to automatically generate a texture. After further reducing mesh complexity by using LOD-techniques, we can map the appropriate texture via classical UV-mapping onto the reduced mesh, such that the detailed color information of the surface is being preserved, even for low polygon surface models.