MSPA-MCS: 3D Scene Digitization - A Novel Invariant Approach for Large-Scale Environment Capture

3D场景数字化：大规模环境捕获的新颖不变方法daniel G. Aliaga, Mireille Boutin， Carl CowenPurdue大学对大型现实环境的模拟是当今计算技术的核心挑战。应用程序数量众多，种类繁多。例如，它将使学生能够虚拟参观著名的历史遗迹，如博物馆、寺庙、战场和遥远的城市；土木工程师捕获建筑物并将其与原始设计或模拟进行比较（例如，在灾难发生前后比较建成模型和模拟模型）；考古学家们几乎保留了复杂的挖掘地点，比如随着时间的推移而演变的沟渠；士兵和消防员在模拟环境中训练；房地产经纪人向买家展示房屋内部；世界各地的人们都可以享受虚拟旅行或多人3D游戏。尽管计算能力和存储空间有了巨大的提高，但目前的获取方法表现得相当糟糕。即使是小场景，它们通常也无法充分捕捉到许多细节。手工创建的模型虽然很流行，但非常耗时，并且渲染图像不能很好地反映现实。另外，基于图像的建模和渲染，产生逼真的图像，但只有在小的和/或漫射环境的背景下，从有限的视点范围看到（例如，QuickTime VR）。同样，专注于重建场景几何形状的方法，如计算机视觉中的重建方法或激光扫描方法，也难以处理复杂的遮挡、镜面和大型数据集。因此，本提案的研究目标是开发捕获和操作大型复杂现实世界场景的视觉丰富的计算机模型所需的算法。本文提出了一种结合场景中包含的几何和光度信息的混合方法来解决这一研究问题。更准确地说，该方法通过密集采样视点空间来捕获3D环境，并使用该冗余数据集提取精确的表面几何模型和场景的反射率属性。这与目前大多数获取稀疏数据集并使用方法插入缺失信息的方法形成对比。这项工作用半自动平台导航、数据过滤和工作集管理等更简单的任务取代了插值。关键是发展高效的数学数据处理技术。该方法的主要研究贡献是将数学和计算机科学的专业知识结合起来，以解决当今计算技术中的一个难题。特别地，该研究利用了一种新的基于李群理论的几何重建方法，该方法是由一个合作pi最近开发的。该方法利用群体作用的一组不变量，消除了三维重构问题中通常包含的一些多余的未知数。正是这些多余的未知量使构造方程变成非线性的。通过删除它们，该方法最终得到一组简单的稀疏线性方程，其中包含最小数量的未知数，可以顺序求解。这使得该项目能够快速而稳健地提取大型数据集的几何（和光度）信息，并重建大型3D环境。拟议的研究将产生超出眼前重建结果的影响。在此之前，研究人员从未接触过如此大规模和密集的环境样本。除了出版物和使所有软件可用之外，该研究项目还将创建一个公共存储库来存储后续研究的模型（例如，历史上重要的地点）。所提出的工作的影响并不是一种更好的捕捉环境的方法，而是一种大胆的新方法，可以显著改变人们对大型环境的计算机模拟的看法。

3D Scene Digitization A Novel Invariant Approach forLarge-Scale Environment CaptureDaniel G. Aliaga, Mireille Boutin, Carl CowenPurdue UniversityThe simulation of large real-world environments is a core challenge of computing technologytoday. Applications are numerous and diverse. For example, it would enable students to pay virtualvisits to famous historical sites such as museums, temples, battlefields, and distant cities; civilengineers to capture buildings and compare them to the original design or to simulations (e.g., tocompare as-built models and simulated models before and after a catastrophe); archeologists tovirtually preserve complex excavation sites such as trenches as they evolve over time; soldiers and firefighters to train in simulated environments; real estate agents to show buyers the interiors of homes;and, people all over the world to enjoy virtual travel or multi-player 3D games.Despite tremendous increases in computational power and storage space, current acquisitionmethods perform quite poorly. Even for small scenes, they usually fail to adequately capture manydetails. Manually created models, although popular, are extremely time-consuming and renderedimages are poor representations of reality. Alternatively, image-based modeling and rendering,produces photorealistic images but only in the context of small and/or diffuse environments seen froma limited range of viewpoints (e.g., QuickTime VR). Similarly, approaches which focus on recreatingthe geometry of the scene such as the reconstruction methods developed in computer vision or thelaser-scanning approaches struggle with complex occlusions, specular surfaces and large data sets.The research objective of this proposal is thus to develop the algorithms needed to capture andmanipulate visually rich computer models of large and complex real-world scenes. The proposalattacks this research problem with a new hybrid method combining both geometric and photometricinformation contained in the scene. More precisely, the approach captures a 3D environment bydensely sampling the space of viewpoints and uses this redundant data set to extract accurate modelsof the surface geometry and the reflectance properties of the scene. This is in contrast with mostcurrent approaches where one acquires a sparse set of data and uses methods to interpolate missinginformation. The work replaces interpolation by the easier tasks of semi-automatic platformnavigation, data filtering, and working-set management. The key is the development of highlyeffective mathematical data processing techniques.The main research contribution of the proposed approach is the merging of expertise from theMathematical and Computer Sciences to solve a difficult problem in computing technology today. Inparticular, the research makes use of a novel geometry reconstruction method based on Lie grouptheory which was recently developed by one of the co-PIs. This method uses a set of invariants of agroup action to eliminate a number of superfluous unknowns normally included in the 3Dreconstruction problem. These superfluous unknowns are exactly the ones that make thereconstruction equations nonlinear. By removing them, the method ends up with a simple set of sparselinear equations involving a minimum number of unknowns which can be solved sequentially. Thisallows the project to quickly and robustly extract the geometric (and photometric) information of largedata sets and reconstruct large 3D environments.The proposed research will have impact beyond the immediate reconstruction results. Neverbefore have researchers had access to such large and dense samplings of environments. Aside frompublications and making all software available, the research project will create a public repository tostore models for subsequent study (e.g., historically significant locations). The impact of the proposedwork is not an incrementally better method for capturing environments, but a bold new approach thatcan significantly change how people think about computer simulation of large environments.