Automatic mesh split-and-merge technique for multiple surface models

In three-dimensional visualization, simulation and modeling applications, surface split-and-merge is a vital technique which creates a complete three-dimensional model that merges individual characterization and properties of multiple surface models. This paper presents an automatic mesh split-and-merge method to handle with the intersection detection, split and mergence processes between multiple surface models. All input models are firstly converted into the uniform triangular representation which retains fine surface features, and the bounding volume hierarchies are built for each model. And then, the hybrid bounding box technique based on the collision detection is used to classify the intersection triangle pairs between models. Furthermore, intersection line segments among these triangle pairs are accurately determined and used to re-mesh models based on the Constrained Delaunay. Finally, input models are split into distinct surface patches which will be merged into a complete surface model based on in-out tests technique. The presented method has been integrated into an automatic model mergence system, and experimental results demonstrate it can effectively and robustly merge multiple surface models.