Virtual Unfolding and Visualization of Papyrus Packages

有关我们祖先的生活和文化的最佳信息来源之一是他们的书面文件。世界各地的博物馆和收藏品中都有大量文献，其中许多已有数百年至数千年的历史，以多种不同语言和各种书写材料书写。片状书写材料，如金属片、羊皮纸、莎草纸和纸，通常被卷起或折叠。物理展开此类文件以显示其书面内容存在永久性损坏珍贵物品的高风险。因此，研究人员已开始使用最先进的断层扫描技术与计算工具相结合来访问此类对象的内容，而无需物理打开对象。对于滚动文档，这已经部分成功，但需要大量工作。在所有书写材料中，纸莎草可能由于其不均匀的双层结构而构成最大的挑战。因此，即使是纸莎草卷轴的虚拟展开，例如赫库兰尼姆纸莎草的研究可能是一个巨大的挑战。更困难的是通过沿多个轴折叠创建的包装的虚拟展开。在前期工作中，我们已经证明，纸莎草纸包的虚拟展开原则上是可能的，但可能会导致巨大的扭曲，从而导致无法破译脚本。此外，使用第一种方法根本无法展开更复杂的包。因此，我们提出了一种新颖的基于模型的方法，其中整个展开过程由交互式可视化引导。虽然这种方法旨在实现最大程度的自动化，但它允许在每个步骤中包含专业知识。这将使纸莎草学家能够高质量地展开非常具有挑战性的（纸莎草纸）卷轴和包装。我们没有根据 3D 图像数据在本地重建纸莎草纸层，而是遵循全局方法，以与真实包完全相同的方式折叠双层网格。然后，通过交互式网格变形和自动网格到图像配准，折叠网格逐渐调整为 3D 图像。对此的一个关键贡献是折叠网格周围图像数据的交互式体积渲染，允许实时响应网格位置和厚度的变化。体积渲染将与网格变形紧密耦合，使纸莎草学家能够通过完全视觉控制实时操纵网格。虽然我们的重点是纸莎草纸包装，但该方法足够通用，可以适用于其他书写材料，也适用于其他应用，例如材料研究中的电池展开或生物学中膜或界面等片状结构的重建。

One of the best sources of information about the life and culture of our ancestors are their written documents. In museums and collections all over the world, there is a wealth of documents, many of them hundreds to several thousand years old, written in many different languages and on a variety of writing materials. Sheet-like writing materials such as metal sheets, parchment, papyri, and paper were often rolled or folded. Physically unfolding such documents to reveal their written content carries a high risk of permanently damaging the precious objects. For this reason, researchers have begun to access the contents of such objects using state-of-the-art tomographic techniques in combination with computational tools without physically opening the objects. In the case of rolled documents, this has been partially successful, but requires extensive work. Of all writing materials, papyrus probably poses the greatest challenge due to its inhomogeneous double-layer structure. Thus, even the virtual unrolling of papyrus scrolls, e.g. of the Herculaneum papyri, can be a great challenge. Still more difficult is the virtual unfolding of packages that were created by folding along several axes. In preliminary work, we have shown that virtual unfolding of papyrus packages is possible in principle, but can lead to large distortions that make deciphering the scripts impossible. Moreover, more complicated packages could not be unfolded at all with this first approach.We therefore propose a novel, model-based approach in which the entire unfolding process is guided by interactive visualization. While this approach aims at the greatest possible degree of automation, it allows expert knowledge to be included in each step. This will enable papyrologists to unfold even very challenging (papyrus) scrolls and packages with high quality. Instead of reconstructing the papyrus layers locally from the 3D image data, we follow a global approach by folding a double-layer mesh in exactly the same way as it was done for the real package. The folded mesh is then gradually adjusted to the 3D image by interactive mesh deformation and automated mesh-to-image registration. A key contribution to this will be an interactive volume rendering of the image data around the folded mesh, allowing real-time response to changes in position and thickness of the mesh. The volume rendering will be tightly coupled to the mesh deformation, allowing papyrologists to manipulate the mesh in real time with full visual control.Although our focus will be on papyrus packages, the approach will be general enough to be adaptable to other writing materials but also to other applications like the unrolling of batteries in materials research or the reconstruction of sheet-like structures such as membranes or interfaces in biology.