Last year, a pair of German artists claimed to have “stolen” scans of a priceless bust of Queen Nefertiti — housed in Berlin’s Neues Museum — using nothing but a Kinect. Now, experts are lining up to call bullshit.
The artists, Nora al-Badri and Jan Nikolai Nelles, say they used “a modified version” of the Kinect to take the unauthorised scans, which they released for free as a protest against what they believe is the museum’s illegal possession of an artefact that should be returned to Egypt.
The thing is… their scans are good. Really good. So good, in fact, that experts in the field of 3D scanning think they’re too good.
As 3D printer Cosmo Wenman recaps on his blog, a number of experts have pointed out that “the Kinect scanner has fundamentally low resolution and accuracy”, and “that even under ideal conditions, it simply cannot acquire data as detailed as what the artists have made available”.
There’s also an especially damning interview Nelles had with Mike Balzer and Chris Kopack from All Things 3D last month, where the limitations of the Kinect’s abilities are made very clear.
Here’s a comparison Wenman put together of the Neues’ Museum’s own hi-res 3D scan and the images released by al-Badri and Nelles:
Image by Cosmo Wenman
“In my opinion, it’s highly unlikely that two independent scans of the bust would match so closely,” he writes. “It seems even less likely that a scan of a replica would be such a close match. I believe the model that the artists released was in fact derived from the Neues Museum’s own scan.”
It’s a mess of a story, in which a mysterious third partner is then blamed for building the scanner and processing the images.
On the bright side, if this has gotten you in the mood to check out some super detailed 3D scans of priceless relics, Wenman’s post also has a nice roundup of other legit images, from 3D scans of the Parthenon friezes to scans of Richard III’s remains.
Photo: Jean-Pierre Lescourret | Getty
Comments
4 responses to “Kinect Involved In Ancient Egypt Kerfuffle ”
The lower image is clearly higher res though.
Not necessarily, a lot of that detail could rather simply be some skillfully added noise and texture to the legitimate scan. The Kinect is an hundred dollar bundle of cameras, it shouldn’t be able to scan anything like this, regardless of modifications.
It’s not exactly higher resolution but the increase in visual quality is due to the fact that the second image is rendered better using better lighting and better rendering methods. The first looks almost like a viewport shot of the scan, a quick and real-time render. The second has better lighting and is using something like ray tracing or global illumination for its lighting. If anything, like the article and experts suggest, the actual scan itself (if you can look past the render quality), is almost identical to the museum one.
I thought I read a thing relatively recently about how some researchers figured out a way to fairly simply increase the res of the Kinect by a factor of about 1000. Let’s see if I can find it.
Here we go:
http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolution-1000-times-1201