Slashdot Mirror


CurvACE Gives Robots a Bug's Eye View

Zothecula writes "Robots are getting down to the size of insects, so it seems only natural that they should be getting insect eyes. A consortium of European researchers has developed the artificial Curved Artificial Compound Eye (CurvACE) which reproduces the architecture of the eyes of insects and other arthropods. The aim isn't just to provide machines with an unnerving bug-eyed stare, but to create a new class of sensors that exploit the wide field of vision and motion detecting properties of the compound eye."

16 comments

  1. Problem by wonkey_monkey · · Score: 1

    Robots are getting down to the size of insects, so it seems only natural that they should be getting insect eyes.

    Robots aren't natural, so no, it doesn't.

    --
    systemd is Roko's Basilisk.
    1. Re:Problem by Impy+the+Impiuos+Imp · · Score: 1

      Does it really matter? Robotic ovipositors aren't natural, either, but as near as I can tell from Internet research, will be very popular.

      --
      (-1: Post disagrees with my already-settled worldview) is not a valid mod option.
    2. Re:Problem by Anonymous Coward · · Score: 0

      Robots are a natural byproduct of humans, so yes

    3. Re:Problem by lxs · · Score: 3, Informative

      The single eye with lens has evolved at least twice on Earth (once in vertebrates and once in arthropods) so it seems like the natural option. Still insects have evolved a different solution for vision. The obvious answer to that is that compound eyes are more practical when your head is tiny.
      I've read somewhere that a mammal style eye on a bee would have to fill its entire head to be useful. A smaller eye would have an aperture only a couple of wavelenghts wide which would render a traditional lens incapable of forming an image.

    4. Re:Problem by ceoyoyo · · Score: 4, Interesting

      The other way around. The maximum resolution of a lens is limited by diffraction. The only way around that limitation is to increase the number of lenses (compound eyes), or increase the size (simple eyes). Increasing the number is a simple solution and works pretty well when you're tiny and so can't have high resolution vision anyway. As you get bigger, the increase-the-size-of-the-lens solution becomes much more efficient, so most bigger organisms have simple eyes. If humans had compound eyes, they would have to be ridiculously large (the size of a house, by one estimation) to give equivalent resolution: http://web.neurobio.arizona.edu/gronenberg/nrsc581/eyedesign/visualacuity.pdf.

    5. Re:Problem by Anonymous Coward · · Score: 0

      Robots aren't natural

      But equivocation on Slashdot is.

  2. Help me! by Anonymous Coward · · Score: 0, Funny

    Help!

  3. Staring sensors by PPH · · Score: 2

    With very wide fields of view and good motion detection. These could put a major dent in UAV operations.

    --
    Have gnu, will travel.
    1. Re:Staring sensors by amiga3D · · Score: 1

      I imagine the intelligence agencies will love them.

  4. not the only one by kermidge · · Score: 2

    There are several similar projects I've read about recently; each in their own way is interesting for their approach and initial impetus, let alone the engineering.

    (One, which I now can't find, was done by a guy at an American university; it was quite large, about the size of a small trashcan and used the guts from existing cameras, While it was a neat project in itself, it was the software he was working on that intrigued me. (My search fu is dead. 45 minutes of using Google to ten pages in didn't show it; searching through two browsers' worth of bookmarks, tags, didn't find it. Also couldn't find his vid on YouTube. Aaargh.))

    Anyway, with respect to the article, if they can get the size down and get useful info for nav and seeing things, it'll open new possibilities. Searching for people in collapsed buildings comes to mind, and there are of course all the surveillance uses.

  5. fish-eye lens by Cyko_01 · · Score: 1

    forgive me if this is a stupid question, but what was wrong with the fish-eye lens? it seems to me the it would be easier to correct for distortion with a fish eyes (constant radius) then having to deal with the Kaleidoscope effect of a bug-eye lens

    1. Re:fish-eye lens by Anonymous Coward · · Score: 0

      Small fast robots flying inside a collapse building do not have time to correct distortions from faulty images. They need real visual data and they need them fast! Otherwise, the danger of collision increases exponentially...

    2. Re:fish-eye lens by X0563511 · · Score: 1

      I can't find the article now, of course... but I recall that there was recent work being done on an imaging technique that uses a compound lens. The lens has no "focal length" in that the entire image is sharp, and the distance of every point is known from the single optic device (no need to use binocular vision (or more, or other sensors) to infer the range, as each input from each part of the lens contributes)

      Computer vision could have a field day with that.

      --
      For large sets, this will be our guide even unto death, for the LORD will work for each type of data it is applied to...
  6. So it's just... by unitron · · Score: 1

    " The aim isn't just to provide machines with an unnerving bug-eyed stare..."

    ...a serendipitous byproduct?

    --

    I see even classic Slashdot is now pretty much unusable on dial up anymore.

  7. Security Cameras and Cars by Flere+Imsaho · · Score: 1

    I can see this being used more to enhance security cameras, rather than robots. Have a couple of bug-eye lenses to monitor 360 deg for movement, then PTZ a high def camera when movement is detected.

    Might also be applicable for self-driving cars?

    --
    It gripped her hand gently. 'Regret is for humans,' it said.