3 ms·
Re: everyone saying the Kinect can't scan that well—remember this paper (http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolution-1000-times-1201 http:/
by derefr 11y ago
Re: everyone saying the Kinect can't scan that well—remember this paper (http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolution-1000-times-1201 http://news.mit.edu/2015/algorithms-boost-3-d-imaging-resolu...) posted on HN a few months back? That research was done by hacking a Kinect.
From the news article:
> 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.
> 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.
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!
ETA: and the 2015 paper isn't even novel. Here's a 2014 paper (PDF: http://ro.uow.edu.au/cgi/viewcontent.cgi?article=5372&context=eispapers http://ro.uow.edu.au/cgi/viewcontent.cgi?article=5372&contex...) 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.
- rsp1984 11y agoNo, it would require three scans -- per viewpoint. But it would require a lot more viewpoints to cover the whole surface area of the object. Also if you look closely at the video footage the capturers use an Xbox 360 Kinect, not an Xbox One Kinect (which is used in the paper you cite) which have entirely different working principles. One other thing that's very fishy: The Xbox 360 Kinect has an infrared projector that's very visible when running. What the activists show in their video is definitely a Kinect that's not running. Finally, keep in mind the paper you quoted is basically lab results. To get it to a level of reliability so it would work as you described a lot more research will have to be done.