Download - Fit for Purpose MCN 2011 mjw
Exploring the Computational Future in Conservation and Heritage Preservation:
Mel Wachowiak and Keats Webb MCN 2011
3D Scanning
Microscopy &3D microscopy
Hyper- & Multispectral
Imaging---and more!
High Dynamic Range Imaging
Reflectance TransformationImaging
GigapixelImaging
Computational imaging the convergence of photography, computer vision, image processing, and computer graphics. Digital processing overcomes limitations of photography and it offers unprecedented opportunities for the enhancement and enrichment of visual media.
In the early days of the Smithsonian, photography had just been invented…and then later we used “film.”
Computational imaging part of integrated group of researchers at the Museum Conservation Institute, located just outside Washington.
• Structured Light
• Laser
• Photogrammetry
Group of Deer, 1941 Paul Manship 1885 -1966
32 1/2 x 27 1/2 x 19 in. (82.6 x 69.8 x 48.2 cm)
Gift of the artist 1963.14.2
Group of Deer Smithsonian American Art Museum
Final merged data, shaded point cloud
MCI contract engineer Anja Schmidt
Excellent hardware, software, and “wetware”
Model made from scan data (rear);
bronze cast awaiting patination
(foreground).
Early daguerreotype 80x70mm
1.8 Gigabyte image from microscopy mosaic.
112 images
All images shown from website screen shots
remember the flaw in the metal plate
It would only require 3000 images to reach resolution of the image at right (0.424mm wide). That’s 30x more than the giga-image.
Research microscope can achieve 5x this magnification, so about 15,000 images needed to reach 1 micrometer or less.
Fit for purpose!
5 stop difference, final 32-bit image; reduced to 8-bit for utility
Paper
Metals
Teeth & Bone
Paintings
Leather
Photo-graphic materials
Stone
RTI project examples
Extended Depth of Field
Z-Axis Scanning
Metrology-based
31 images in this stack
4mm trade bead from African bag
Before and after cleaning
Before and after cleaning
• While we can, when should we?
• n-dimensional: time and space, etc.
• Fit for purpose requires understanding of methods and resources.
• Who will use this data in 165 years?
If you can imagine it, it can be done.