Towards Globally Optimal Normal Orientations for Large Point Clouds
Citations Over TimeTop 15% of 2016 papers
Abstract
Abstract Various processing algorithms on point set surfaces rely on consistently oriented normals (e.g. Poisson surface reconstruction). While several approaches exist for the calculation of normal directions, in most cases, their orientation has to be determined in a subsequent step. This paper generalizes propagation‐based approaches by reformulating the task as a graph‐based energy minimization problem. By applying global solvers, we can achieve more consistent orientations than simple greedy optimizations. Furthermore, we present a streaming‐based framework for orienting large point clouds. This framework orients patches locally and generates a globally consistent patch orientation on a reduced neighbour graph, which achieves similar quality to orienting the full graph.
Related Papers
- Коммуникaтивно- прaгмaтический aнaлиз дипломaтических бумaг (нa основе вербaльных нот)(2018)
- FEATURES AND DIFFERENCES OF ADEQUATE AND EQUIVALENT TRANSLATION(2018)
- चितलवाना पंचायत समिति में मानव गरीबी सूचकांक - 2016 ( à¤à¤• गà¥à¤°à¤¾à¤® सà¥à¤¤à¤°à¥€à¤¯ à¤à¥Œà¤—ोलिक अधà¥à¤¯à¤¯à¤¨ )(2017)
- → ФОРМИРОВAНИЕ ГОТОВНОСТИ БУДУЩИХ ПЕДAГОГОВ К ОРГAНИЗAЦИИ РAБОТЫ ПО РAЗВИТИЮ ВAЛЕОЛОГИЧЕСКОЙ КУЛЬТУРЫ ШКОЛЬНИКОВ(2023)