EAGER: Hiding structure in geometric data exchange formats

点击翻译按钮获取中文摘要

A fundamental algorithmic/software issue in 3D printing, visualization, and geometric computing is incompatible file-exchange formats. For instance, 3D printing data is usually stored in a format called STL, but there are numerous geometric tools which use other (incompatible) formats (some well-known extensions being: STH, RPI, and STEP, to name just a few). This project will pursue a novel direction to improve compatibility and efficiency in geometric computing: restructuring files, in their native format, without translating them into a new format. This will enable faster checking of 3D models and processing. A key novelty is that this work will use existing file formats rather than create a new one. Potential impact include faster modeling of complex geometric data and real-time remote collaboration on complex designs, such as mechanical parts or architectural data. The restructuring and preservation of native file formats is technically quite involved, and leads to several interesting research directions: (1) How much additional geometric information can be ``hidden'' in a file simply by reordering the data? (2) What is the most useful geometric information that should be ``hidden'', (3) What is the most effective way to pack information for an application while still preserving the file format? A key example to keep in mind is how STL redundantly stores vertices of facets, and that this enables one to introduce a new ordering of the redundant data without disrupting the underlying file format. Answering these questions can also help support streaming of geometric data.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.