A Dynamic Hierarchy Based on Adaptive Segmentation

According to surface characters of the models, the ASLOD (Adaptive Segmentation Level of Detail) approach is presented in the paper. This algorithm is an efficient multi-resolution triangulation framework based on the half-edge triangle mesh data structure. It takes advantages of the sharp features of the models to perform adaptive segmentation and simplification, with an approach ensuring the segmentation area simplified independently without creating tears or cracks. The natural hierarchy provided by this algorithm makes it particularly suited for applications such as progressive transmission and LOD hierarchy construction. The results show that the method is highly efficient in terms of both space and time cost, and it can render large models with important model features well preserved.