Spline-based feature curves from point-sampled geometry

Abstract

Defi ning sharp fea tures algorithm processes these points to construct arc-length parame teri zed splin e curves fi t using an iterative refin eme nt method, ali gnin g smooth and continuous curves through the feature points. We demonstrate the benefits of our method with three applications: surface segme ntation, surface meshing and point-based compression. in a 3D model fac ilitates a better understanding of the surface and aids geometric processing and graphics applications, such as reconstruction , filtcring, simplificati on, revcrsc engineering, visualization, and non-photo realism. We present J. Danie ls II (~). T. Ochotta• L. K. Ha • c.T. Silva Scientifi c Computing and Tmaging Tnstitute Universi ty of Utah a robust method that identifies sharp features in a point-based model by returning a set of smooth spline curves aligned along the edges.

Sharp feature preserving MLS surface reconstruction based on local feature line approximations

“…and how to chose this point? These questions are partially answered in Daniels et al [7]. However, the critical mass necessary for [7] is not given here.…”

“…These questions are partially answered in Daniels et al [7]. However, the critical mass necessary for [7] is not given here. The solution we propose is consistent to the rest of the paper, since using again the Gauss-map clustering technique.…”

“…Although the resulting surface is of a good quality, the outlier search and the refitting of the smooth surface parts are quite tricky to implement. The method also has problems with jagged edges as pointed out in [6,7] and needs a quite dense sampling. Our approach is inspired by the partitioning of the local neighborhood to reconstruct sharp features, although the way how we achieve this is completely different.…”

“…Sharp corners and edges can be reconstructed by selecting appropriate shape functions. Daniels et al [6,7] build up on Fleishmans approach [9] to project points onto the features of the dataset. A covariance analysis for smoothing and growing polylines of the detected feature points leads to a global approximation of feature lines.…”

“…• The method is much more simple to implement than statistical methods [9,6,7,21] where many parameters have to be fine-tuned.…”