One possibility is that there were no 'hacker' types. The scan might have just been given to them by someone sympathetic to their cause, and the easiest way to avoid investigation while still getting press is to carry in an unplugged Kinect and claim to have made the scan yourself.
I got side tracked over the discussion of Germany "not being amused" by this.<p>Wait, so this bust was made by Egyptians over 3000 years ago and the Germans are not happy someone made a scan of it!? It's almost disgusting that anyone can lay claim to this work of art! It should "belong" to ALL of us at this point.
Could it be a scan of a $39[1] or a $2,990[2] replica?<p>1: <a href="https://www.ancientsculpturegallery.com/nefertiti-small-sculpture.html" rel="nofollow">https://www.ancientsculpturegallery.com/nefertiti-small-scul...</a><p>2: <a href="https://www.ancientsculpturegallery.com/egyptian-queen-nefertiti-bust-sculpture-identical-museum-reproduction.html" rel="nofollow">https://www.ancientsculpturegallery.com/egyptian-queen-nefer...</a>
The long and short of it is this: if they got a scan of that detail from a Kinect, of an object under glass then they better release a SIGGRAPH white-paper because what they accomplished was better than the state of the art.<p>Of course there's a much more likely explanation.
I spent years working on algorithms for processing 3D scanner data, and the claim that you could get a scan of that resolution and quality - surreptitiously, through a glass enclosure, and with a Kinect - is absurd.<p>Most likely it's a scan of a replica created under carefully controlled conditions, or they were able to acquire a scan created by the museum.
Re: everyone saying the Kinect can't scan that well—remember this paper (<a href="http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolution-1000-times-1201" rel="nofollow">http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolu...</a>) posted on HN a few months back? That research was done by hacking a Kinect.<p>From the news article:<p>> The researchers’ experimental setup consisted of a Microsoft Kinect — which gauges depth using reflection time — with an ordinary polarizing photographic lens placed in front of its camera. In each experiment, the researchers took three photos of an object, rotating the polarizing filter each time, and their algorithms compared the light intensities of the resulting images.<p>> On its own, at a distance of several meters, the Kinect can resolve physical features as small as a centimeter or so across. But with the addition of the polarization information, the researchers’ system could resolve features in the range of tens of micrometers, or one-thousandth the size.<p>Not a hard hack to accomplish! It's just a piece of plastic, and then a compressed sensing algorithm (which you can run at your leisure on your scan data after leaving the building) to merge the UV maps. And it doesn't even require a thousand scans or anything—just three!<p>ETA: and the 2015 paper isn't even novel. Here's a 2014 paper (PDF: <a href="http://ro.uow.edu.au/cgi/viewcontent.cgi?article=5372&context=eispapers" rel="nofollow">http://ro.uow.edu.au/cgi/viewcontent.cgi?article=5372&contex...</a>) on using multipolarized compressive sensing to enhance the resolution on through-wall radar. (I assume to target drone strikes or something.) This technique has been in the wild for a while. I assume archivists—those who would most benefit—are well-aware of it.
> stolen scan<p>Erm, even if it's scanned, it's a scan of an object that is pretty much in public domain. Not sure the use of the word "stolen" is appropriate here (as usual) - it's more like they got access to it.
On the subject of ancient art, CBC radio did a interesting series on who should get to say where and how ancient art gets kept and displayed. It's a two-part series and covers a lot of ground ...<p><a href="http://www.cbc.ca/radio/ideas/who-owns-ancient-art-part-1-1.3106590" rel="nofollow">http://www.cbc.ca/radio/ideas/who-owns-ancient-art-part-1-1....</a>
<a href="http://www.cbc.ca/radio/ideas/who-owns-ancient-art-part-2-1.3119029" rel="nofollow">http://www.cbc.ca/radio/ideas/who-owns-ancient-art-part-2-1....</a>
I agree, it does seem suspicious. The other thing I noticed is that the top of the hat/headdress is scanned accurately, which would be difficult unless you held the scanner in an obvious manner above the statue.
Interesting stuff. As a once kinect enthusiast I was completely blown away by the quality of this thing. Normally, the kinect is a pita to deal with, is very touchy, and gives a faily low quality/res return. I just assumed they were using some fancy image stabilization and some fuzzy logic/hand cleanup to make it not look terrible. Or that the Xbox One kinect was just much higher resolution than the original, but a quick google search tells me the difference is negligible. Hell, if the kinect could do resolutions at this level natively, everyone would buy one as it would exceed 3D laser scanners costing thousands, if not tens of thousands of dollars.<p>The hacker theory is interesting, but I wonder if someone with access to this statue during off-hours performed this with a proper laser scanner and the kinect story is cover to make sure he doesn't get caught. Sounds like someone might be in trouble now. Of all the countries to have an intellectual freedom debate/trial, it seems that politically unstable Egypt would be one of the worst for any positive outcome here.<p>Its all a shame. We should have good 3D scans of every museum item of note by now and available for free for the public. How can we sell this concept to the boards, trustees, curators, and bureaucrats who control these items? Wasn't there a big controversy when high-quality color film came out way back when about publishing photos of art pieces and museum pieces? Seems like we're destined to re-invent this conflict over and over as technology advances.<p>edit: This bust is permanently housed in Germany, I just assumed it was on temporary loan. Thanks for the correction below.
The second version of the Kinect uses time-of-flight sensing rather than the projected grid of points, and is considerably more accurate.<p>Whether or not this makes the scanning more plausible, I am not sure. However, some of the 3D models generated with the Kinect 2 [1] do seem to contain a decent amount of detail.<p>[1] <a href="http://zugara.com/wp-content/uploads/Kinect-1-vs-Kinect-2-Visual-Comparison.jpg" rel="nofollow">http://zugara.com/wp-content/uploads/Kinect-1-vs-Kinect-2-Vi...</a>
This wouldn't be the first time artists taking credit to do one thing while actually doing something else.
After all the truth is probably way too boring for it to be considered art.
It's a great stunt if someone at the museum "leaked" this. If there is a person like that working there, his/her people skills are beyond believable/demagogic...<p>Makes me think of the book <a href="https://en.m.wikipedia.org/wiki/Look_Who%27s_Back" rel="nofollow">https://en.m.wikipedia.org/wiki/Look_Who%27s_Back</a>
Agree with the poster. We did some scanning experiments to do busts of ourselves with a 3d printer as a first test with various tools. It's pretty hard to scan a human and still get a lot of detail. Not sure how much time they had and if scanning a bust is easier (it's under glass though) but the quality seems too good at first glance.
> What do you think? Will we ever know?<p>Of course we will know. If the data is from a scan the museum made, there exists a way to compare the two data sets and give a certainty one came from the other.<p>That is the nature of entanglement, after all.
I too go for access to the museum's 3D scan data somehow. That being said, here are some thoughts on the Kinect 1 being used to produce a scan given the display glass, the power requirements, and resolution of the Kinect1.
The Kinect1's power requirement goes way down once you cut the motor's power out. Second, perhaps the Kinect was used with the intention of helping the photogrammetry out, by giving precise distances to the glass for each shot taken by the Kinect1, and then secondary photos from cell phones could be correlated back at home. I doubt the use of a polarizing filter, since it would not lend much to this method. In addition, you can double the resolution of the Kinect1 by not streaming at a high rate. I believe from 640x480 VGA to 1280x1024 at a lower frame rate. If the display glass were not there, they could get at best 1.3mm per pixel at those distances. But it was there, so I vote for a file provided to them. I am just presenting the technical possibility of augmenting the Kinect1 with some hacks and cell phone photos to get close. I do think in any case mesh cleanup would allow the 1.3mm per pixel to be sub-surfaced to 0.01mm vs. 1.3mm stated above without any major infidelity to the original sans micro-scratches and pores on the original surface.
I worked for 2 years using ultrasound, photos and video to digitize objects before the Kinect1 appeared, so I know what can be done to achieve reproducibility of a scanned object. NASA had researched 'shape from shading' algorithms in the 70s and 80s to try and get more data from their mono B&W satellite imagery. I had implemented these in my own products, and no, they are not simple height maps based on a greyscale image. They take into account the light source's position to extrapolate relief data on monotonous toned objects like asteroids and unpainted sculptures.
Power and laptop requirements are the easiest to solve here, she has a big trenchcoat. You can easily hide enough battery power to power kinect. And at the end of the day kinect just outputs signal so why can't you have something small like RPi or phone to take in the data and send it to a more powerful computer for processing.
Most likely hypothesis to me is that they scanned the object, produced a low quality scan, then somebody rebuilt the scan using surface primitives from the low quality scan. The result is not accurate, but is very precise.
Of course the Kinect does not offer such precision. But what about other pro hand held scanner?<p>HandySCAN, Polhemus, Artec, all 3D scanner brands with a very good resolution.<p>But then again it seems very unlikely that they took the time to scan an object without being noticed.
there is a lot of press around this work but seems derivative of material speculation ISIS: <a href="https://www.youtube.com/watch?v=NL6e-kEkur0" rel="nofollow">https://www.youtube.com/watch?v=NL6e-kEkur0</a>