[FoRK] It's Time to Start 3D Scanning the World
sdw at lig.net
Wed Jan 11 07:22:54 PST 2012
No special hardware is needed to get very good 3D scans, once we get a little more CPU and improve the algorithms just slightly.
Stereo vision, using just a single camera moved around a bit, is plenty to get pretty good data.
Imagine we bootstrap robots with just the basics, but great vision, memory, CPU, and networking. We then crowd source teaching of
shapes, objects, paths, patterns, jobs, etc. And we would obviously have them share everything they learned on the robosocial
RoboBook and RoboWiki.
What we really want to teach is basically the same thing that humans use to generalize and compress: prototypical objects that can
be fitted to a wide range of instances. You can do that with the many-examples method (which works best now for many things) or
some combination of logical description and component recognition (probably better for things like chairs).
However, great databases of 3D objects would be great.
On 1/11/12 6:59 AM, Eugen Leitl wrote:
> It's Time to Start 3D Scanning the World
> POSTED BY: Erin Rapacki / Tue, January 10, 2012
> This is a guest post. The views expressed in this article are solely those of
> the blogger and do not represent positions of Automaton, IEEE Spectrum, or
> the IEEE.
> matterport scan
More information about the FoRK