1. National Center for Supercomputing Applications University of Illinois at Urbana-Champaign Date: January 21 st, 2009 Appraisal of 3D Data Conversions and Visualization Software Packages Peter Bajcsy, PhD - Research Scientist, NCSA - Adjunct Assistant Professor ECE & CS at UIUC - Senior Associate Director Center for Humanities, Social Sciences and Arts (CHASS), Illinois Informatics Institute (I3), UIUC

2. Acknowledgement This research was partially supported by a National Archive and Records Administration (NARA) supplement to NSF PACI cooperative agreement CA #SCI- 9619019. The views and conclusions contained in this document are those of the authors and should not be interpreted as representing the official policies, either expressed or implied, of the National Archive and Records Administration, or the U.S. government. Contributions by: Peter Bajcsy, Kenton McHenry, Rob Kooper, Michal Ondrejcek, William McFadden, Sang- Chul Lee, David Clutter and Alex Yahja Imaginations unbound

3. Outline Introduction Problem Formulation Approach Prototype Solution -Polyglot Conclusions

4. Why Appraisal of 3D Digital Data? Preservation of cultural heritage Museums, Libraries Preservation of engineering drawings Companies: Boeing, John Deere, GM Government agencies: NASA, DoD Preservation of architectural designs Construction and architecture firms Preservation of simulation results Companies: Caterpillar, Ford Government agencies: EPA, FEMA, DoD Preservation of medical data Companies: GM Government agencies: DoD Imaginations unbound

5. How To Share 3D Data? Many File Formats Complex & Proprietary Standards (adoption & implementation) Packaging One file Many files Embedded in a container(e.g., PDF, HDF) Imaginations unbound

6. How to Manage 3D File Formats?

7. How to Compare and Relate 3D Data? Example Data: Torpedo Weapon Retriever 841 784 existing 2D image drawings and 73 3D CAD models How to establish relationships among the 3D CAD models and 2D image drawings during a product lifecycle? Imaginations unbound Hypothetical Distribution of 3D CAD models for TWR 841

8. How to Design Scalable Appraisal Approaches? Scalability of appraisals with increasing heterogeneity of information, dimensionality of data and volume of electronic records How should appraisal process change as 3D data is added to file containers? How should appraisal process change as 3D+time, 2D+spectrum, 3D+time+spectrum, nD, … How should appraisal operations be designed to accommodate growing volume of electronic records? Imaginations unbound

9. Problem Formulation Imaginations unbound

10. Basic Archival Appraisal Questions Are the 3D formats well formed? Can we identify a minimal set of information to preserve? Is there an optimal format to convert to? Can we quantify 3D noise introduced during conversions? Can we quantify differences in renderings?

11. Current Focus on 3D Data Conversions 3D file format conversions are inevitable since (a) various software packages support importing and exporting different file formats, (b) many file formats are retired or deprecated over time, (c) the cost of preserving the existing large number of file formats is too high. There is a need to understand the impact of 3D data conversions and the quality of the used conversion and visualization software with respect to information preservation. Imaginations unbound

12. A Subset of Conversion Related Problems Problem #1: Appraisal of conversion paths for a given pair of source and target file formats Problem #2: Appraisal of conversion software Problem #3: Decision support for a target file format selection Problem #4: Decision support for information preservation metrics Imaginations unbound

13. Approach Imaginations unbound

14. Methodology Design a graph model of the conversion space Assign characteristics to nodes and edges of the graph Find all paths that connect source and target nodes of the graph Evaluate dissimilarity metrics and computational costs Analyze and appraise accordingly Optimally process for preservation Imaginations unbound

15. Graph Model of Conversion Space Imaginations unbound

16. Graph Nodes: 3D File Format Survey From www.turbosquid.com ~ 3D market placewww.turbosquid.com

17. Graph Oriented Edges: 3D Software Survey Adobe 3D Reviewer

18. Dissimilarity Metrics (Work in Progress) 2D histograms, e.g., spin images, 3D light fields, e.g., visual hull, metadata, e.g., header information and material properties, appearance, e.g., screen snapshots, …... Imaginations unbound Spin Images [Johnson, PAMI ‘99] - Construct 2D histograms for each vertex: -  : distance to plane represented by vertex normal -  : distance to vertex on this plane - Cluster histograms for each model - Compare using Euclidean distance

19. Observations: Conversion of X3D to STEP to X3D STEP WRL X3D WRL Software: X3dToVrml97 X3D Software: A3D Reviewer Software: A3D Reviewer Software: Vrml97ToX3d Nothing!

20. Observations: STP to X3D to STP X3D WRL STP WRL A3D Reviewer STP Vrml97ToX3d X3dToVrml97A3D Reviewer

21. Prototype System: Polyglot Imaginations unbound

22. Current Components of Polyglot Input/Output Graph Around 140 file formats (graph nodes) Seven open source software packages (Blender 2.46, Cyberware PlyTool (iplytool-071024), K3D 0.6.7, VTK 5.2.1, Wings 3D 0.99.02, NIST's Vrm97ToX3d v1.0, NIST's x3dv2) One commercial package (Adobe 3D Reviewer) Conversion Service Scripted execution Visualization Service Imaginations unbound

23. Input/Output Graphs Adobe 3D Reviewer

24. Several converters Input/Output Graphs

25. Shortest Path From Input/Output Graphs Shortest Path

26. Conversion Service Challenges with conversion software: No application programming interfaces (API) for import and export functionality in software packages Software packages are closed source Software packages are user interface driven Approach: The conversion is automated by using the Windows AutoHotKey scripting language http://www.autohotkey.com/ Limitations: Only one conversion can be executed at the same time with one software package due to license issues Imaginations unbound

27. Polyglot as a Web Service: Upload/Convert/Download Imaginations unbound View after downloading using a Desktop Tool

28. Polyglot as a Web Service: Find Mapping Specify Source and Target File Format Imaginations unbound

29. Polyglot as a Web Service: Display/Interact/Visually Compare Imaginations unbound

30. Conclusions Imaginations unbound

31. Current Status For us, Polyglot is a research prototype that serves as a foundation for addressing basic archival appraisal questions Noise introduced by conversions Quality control Appraisal scalability There is a value in our better understanding of computational and storage costs associated with preserving 3D data Our current work also enables addressing the questions introduced at the beginning, for example, discovering relationships and 3D comparisons (work in progress) Imaginations unbound

32. Collaboration Areas with PRONOM/DROID/JHOVE Teams 2D Image (JPEG) 3D Data (IGS) Imaginations unbound

33. Questions Project URL: http://isda.ncsa.uiuc.edu/NARA/index.html and http://isda.ncsa.uiuc.edu/CompTradeoffs/ http://isda.ncsa.uiuc.edu/NARA/index.html and http://isda.ncsa.uiuc.edu/CompTradeoffs/ Publications – see our URL at http://isda.ncsa.uiuc.edu/publications http://isda.ncsa.uiuc.edu/publications Peter Bajcsy; email: pbajcsy@ncsa.uiuc.edupbajcsy@ncsa.uiuc.edu