Unifying NURBS and subdivision: Extracting sparse shape descriptions using NURBS-compatible subdivision

我们将研究在工程（汽车，飞机，船舶）和娱乐（动画和特效）中改善3D设计师工作流程的方法。我们从我们最近的工作开始，调和了3D模型的两种主要表示：NURBS，用于工程和娱乐，以及细分表面，主要用于娱乐。我们的工作为工程应用的严格要求带来了细分方法的灵活性，而不必放弃NURBS方法的显着优势。这是30年来的研究目标。我们最近的论文是实现这一目标的第一篇工作（具有非凡点的NURBS：高度非均匀细分表面，发表在著名的ACM SIGGRAPH会议和ACM期刊Transactions on Graphics上）。这笔拨款以两种相互联系的方式推进这项工作。Strand 1研究了新方法如何加速3D模型的捕获。许多工程和娱乐模型仍然用粘土制作。必须捕获粘土模型，然后使用NURBS或细分将其转换为数字表示。这第一股将着眼于我们可以捕获一个粘土模型的方式，并产生一个良好的内部表示，可用于进一步操纵模型。我们的新建模方法将在这里受益，因为它允许我们使用细分的灵活性，同时仍然与工程友好的NURBS方法兼容。Strand 2扩展了我们新方法的功能。我们将研究各种可以改进的方法，为3D设计师提供更丰富的建模工具。这些改进将反馈到Strand 1中，而Strand 1中的经验将为我们在Strand 2中研究的特定改进提供信息。如果我们成功了，我们可以期待我们的方法被主要的计算机辅助设计系统所采用。这将带来3D设计工作流程的改进，因此汽车，飞机和动画人物可以比以前更快，更灵活地建模。

We will investigate methods to improve the work-flow of 3D designers, in engineering (cars, aircraft, ships) and entertainment (animation and special effects). We begin with our recent work that reconciled the two principal representations of 3D models: NURBS, used in both engineering and entertainment, and subdivision surfaces, used principally in entertainment. Our work brings the flexibility of the subdivision approach to the rigorous requirements of the engineering applications without having to forgo the significant advantages of the NURBS approach. This has been a research goal for thirty years. Our recent paper is the first piece of work to achieve this ( NURBS with extraordinary points: high-degree non-uniform subdivision surfaces published at the prestigious ACM SIGGRAPH conference and in the ACM journal Transactions on Graphics ).This grant carries forward the work in two inter-connected strands.Strand 1 investigates ways in which the new method can speed the capture of 3D models. Many engineering and entertainment models are still made using clay. The clay model must be captured and then converted into a numerical representation, using either NURBS or subdivision. This first strand will look at ways in which we can capture a clay model and produce a good internal representation that can be used for further manipulation of the model. Our new modelling method will be of benefit here, as it allows us to use the flexibility of subdivision, while still being compatible with the engineering-friendly NURBS method.Strand 2 extends the capabilities of our new method. We will investigate various ways in which it can be improved, to provide a richer modelling tool for 3D designers. The improvements will feed across into Strand 1, while the experience in Strand 1 will inform the particular improvements that we investigate in Strand 2.If we are successful, we can expect our methods to be taken up in the main computer-aided design systems. This will bring improvements in 3D design work flow, so that cars, aircraft, and animated characters can be modelled more quickly and more flexibly than before.