ACRL announces the publication of 3D Data Creation to Curation: Community Standards for 3D Data Preservation, edited by Jennifer Moore, Adam Rountrey, and Hannah Scates Kettler, which captures best practices for 3D data preservation, management, access, and more.
Learn more about 3D Data Creation to Curation in this Introduction from the editors, licensed under CC BY-NC.
Context for This Work
There has been rapid growth in the production and usage of 3D data over the last decade, yet the preservation[*] of these data has lagged behind to the detriment of scholarship and innovation. While the need for digital 3D data preservation is widely recognized, the ongoing development of 3D data creation processes and the evolving usage of content still present many open-ended questions about how to ensure the stability and durability of this data type. Creators, curators, and users of 3D datasets are disadvantaged by the lack of shared guidelines, practices, and standards.[1] This volume, which includes surveys of current practices, recommendations for implementation of standards, and identification of areas in which further development is required, is a result of the efforts of a large practicing community coming together under the Community Standards for 3D Data Preservation (CS3DP) initiative to move toward establishment of standards.[2]
The goal of this work is to identify the broad, shared preservation needs of the whole community, and it is viewed as essential to use a collaborative approach for standards development that promotes individual investment and broad adoption. The authorship of the chapters recognizes those who worked to discuss particular aspects of preservation in detail, but throughout the process of development, the entire community has been engaged, shaping the content to meet needs across a diverse base of stakeholders.
The Audience
The CS3DP project was established to bring together people with diverse backgrounds and experiences with digital 3D data (henceforth, 3D data)[†] to examine the current practices in 3D data documentation, dissemination, and preservation and to make recommendations for standardization that could bring broad adoption and benefits. For example, efforts that seek to preserve physical objects through digitization are not worthwhile without means of preserving digital products, but this is one of many contexts where 3D preservation is beneficial. The people who make up the community have backgrounds in art, architecture, natural history, information science, medicine, archaeology, and law, and there are representatives from academic institutions, nonprofits, and commercial industries mostly based in the United States; they also have experience with a diversity of approaches.
Given the potential scope of the work that this book attempts to cover and complexities arising from different legal frameworks in other countries, it is written for a primarily US audience, although much of what is covered will no doubt be applicable to those outside the US. This work likely has relevance beyond the borders, and the authors did attempt to make reference to related work going on in other areas of the world. It should also be noted that the discussions tended to focus on 3D applications in academic research, cultural heritage, and education, but the resulting material will still be useful to those operating outside of educational contexts.
These pages are intended to be used by people with varying amounts of 3D experience, from novice to seasoned practitioner, and are also intended for people in 3D data preservation support roles, who may or may not be involved in the creation of the data, yet may be tasked with curating, migrating, and sustaining access to these data long-term. This last task, sustaining access to these data, is the crux of the problem. Regardless of our intent, whether digitizing a physical space or object, representing imaginative spaces, or creating 3D for entertainment, without some kind of access to these data, they likely will escape preservation. 3D models can be expensive consumables or flashy ephemera, yet in some cases, such as with entities like CyArk that are meant to empower the collective preservation of cultural heritage material in 3D, that ethos can be undermined by a lack of perpetual access and long-term preservation.[3] Access and preservation, and the steps in between, are inseparable from each other. Though this work is indeed focused on preservation, the pages within reflect the entire life cycle of 3D data creation and maintenance, underpinned by concerns of access by various user groups.
From Creation to Preservation
The process of preserving 3D data increases FAIRness (findability, accessibility, interoperability, reuse). CS3DP working groups approached the problem through their respective lenses: best practices planning, metadata and documentation, long-term repository storage and management, articulation of rights and ownership, and restrictions and access. These topics are distinct and interconnected; they make up the framework for this volume and will be presented as independent chapters, which reference and leverage their companion chapters.
To begin to contextualize the problem, there are two main branches of 3D model generation: reality capture and manual modeling. Reality capture means the creator has a physical thing that they would like to replicate using a camera or scanner. Within the reality capture branch are two more major distinctions: whether the model is volumetric or a surface capture. Volumetric data are data that include measurements or other values in a 3D array or grid. A common method that produces volumetric data is computed tomography (CT) scanning. Surface-based captures aim to digitally reproduce the shape of the object in 3D but do not collect density information. Manual modeling also often creates a 3D surface object, but not via a digital capture. Common manual creations are results of creative modeling and reference-based modeling. Models can also be created by combining multiple methods. Specific methods or modes of creation will be described in the next section.
In the best case scenario, the life cycle of 3D data begins, as much data creation does, with planning. Among the questions that need addressing in the planning stage are what is the purpose of generating a 3D object, what is the origin information that will contribute to its creation, and what mode of creation best fits the purpose and origin. Often the purpose and originating information guide the creation tools and methods. Chapter 2, “Best Practices for 3D Data Preservation,” will articulate how careful planning can impact the preservation of data for various modes of creation. That chapter will plant many seeds that will flourish throughout the rest of the volume.
Modalities Represented in the Chapters
There are two broad digital data types covered in this book: (1) data that may include 3D points, edges, and faces, such as a polygonal mesh representing the surface of, for example, a scanned statue, and (2) volumetric, or voxel, data, which are a 3D array or grid with values assigned to cells in the grid (e.g., CT scan data). There are a variety of ways in which data in these two types are produced. Some of the most common methods are listed below.
3D point/mesh data:
- Photogrammetry: The extraction of three-dimensional measurements from two-dimensional data (i.e., images). Developments in GPU-based processing allow rapid reconstruction of 3D surface meshes from sets of conventional photographs of a physical object or environment. The mesh output from this technique may be enhanced by color information at vertices (i.e., vertex color) or an associated 2D image representing surface color, which is mapped to the mesh (texture map).
- Laser scanning: The process of recording precise three-dimensional information about a real-world object or environment by rapidly sampling or scanning an object’s surface with lasers. The information is often returned to the user as a dense collection of precisely located x,y,z coordinates referred to as a point cloud. Laser scanning devices may use a time-of-flight method, a phase method, or a triangulation method. The point cloud or mesh output from this technique may be enhanced by color information at points/vertices (i.e., vertex color) or an associated 2D image representing surface color, which is mapped to the mesh (texture map).
- Structured light: Method of 3D capture that relies on the distortion of projected light to calculate surface form. A known pattern (often a grid or horizontal lines) of light projected onto a surface appears distorted from perspectives other than that of the projector. This distortion can be used for geometric reconstruction of the surface shape. The mesh output from this technique may be enhanced by color information at points/vertices (i.e., vertex color) or an associated 2D image representing surface color, which is mapped to the mesh (texture map).
- Bibliography/sources-based modeling: A method of model production based on documents, reference photographs, or other sources of information about a real-world object or place. Models are often created in a CAD (computer-aided design) software system.
- Creative modeling: A method of model production in which the user designs a 3D object or environment based on creative vision.
Volumetric data:
- CT scanning: Also known as a computed tomography scan, and formerly known as a computerized axial tomography scan or CAT scan. This method makes use of computer-processed combinations of many X-ray measurements taken from different angles to produce cross-sectional (tomographic) images (virtual “slices”) of a scanned object, allowing the user to see inside the object without cutting. Other medical imaging methods (e.g., MRI) produce similar volumetric data based on different properties of the object.
- Voxel art: A method of modeling in which objects are represented by many 3D cubes that may vary in color (e.g., “Minecraft style”).
Multimodal modeling:
- A 3D model resulting from a combination of methods (modes).
While dividing these data types by representation (points and meshes versus volumetric data) is useful for many parts of 3D data preservation planning, there are times when it is more useful to separate data by whether or not they are a faithful attempt at recording the geometry and characteristics of a measurable (at the time of capture) real-world object or environment, a faithful attempt at recreating geometry and characteristics of a real but nonmeasurable object or environment, or a creative output in which expression is more important than representation of a real-world object or environment. In this book, we will refer to reality-capture models, sources-based models, and creative models respectively to differentiate these when appropriate.
Many of the methods mentioned above will be covered in more detail in chapter 2, “Best Practices for 3D Data Preservation,” and referenced throughout the book. As much as the authors could, those different modalities are acknowledged and considered, and the discussions and recommendations are made with the myriad pathways of 3D data creation in mind. We chose to focus mostly on commonly used methods that do not have well-established preservation standards, so some types of data, such as medical CT and 3D GIS data, are not discussed in detail.
[*] All terms in bold type are defined in the glossary.
[†] The glossary included in this book was developed collaboratively out of necessity. We quickly realized that in order to understand each other as we spoke about our 3D data work across the various disciplines and modalities, we needed a set of common terms that could be used across these boundaries.
[1]. Community Standards for 3D Data Preservation (CS3DP), “CS3DP 2017 Community Survey,” accessed January 10, 2020, https://osf.io/tcn6h/.
[2]. University of Iowa, “Community Standards for 3D Preservation (CS3DP),” accessed May 4, 2020, https://ir.uiowa.edu/cs3dp/.
[3]. CyArk home page, accessed January 15, 2020, https://www.cyark.org/.