Follow Slashdot blog updates by subscribing to our blog RSS feed

 


Forgot your password?
Close
typodupeerror
Science

Camouflage in Motion 57

Posted by michael from the predator dept.
Adrian writes "Remember Jurassic Park, where Goldbloom stood really still and the T-Rex couldn't see him? Well, there might be a better way. Scientists have found that dragonflies can dissappear by keeping their image on your retina in the same place, even if you move. How they manage it still has them puzzled... ;)"
This discussion has been archived. No new comments can be posted.

Camouflage in Motion More

Camouflage in Motion

Comments Filter:

  • Hm (Score:5, Funny)

    by cjpez ( 148000 ) on Friday June 06, 2003 @11:27AM (#6132262) Homepage Journal
    What an incredibly hollow article. "We've used some technical majiggers to look at some stuff and wow! Look what we came up with!" It's a good thing there wasn't any actual details in there, it may have been interesting.

    • Re:Hm (Score:5, Informative)

      by Cy Guy ( 56083 ) * on Friday June 06, 2003 @12:39PM (#6132962) Homepage Journal
      Discovery.com provides a few more details [discovery.com] but since the scientists themselves are still baffled, I don't think we will find any lengthy explanations of the phenomenom except perhaps by reading the article in Nature itself which is not available except by subscription.

      The thing new in the Discovery article I found significant was that they performed the movements with "millimetric" precision.

      I wonder if the dragonfly's 3 foot long ancestors were also capable of such precision, or whether the need to remain so precise led to their reduced current size.

      • Re:Hm (Score:2)

        by cjpez ( 148000 )
        Yeah, that Discovery article is a bit more informative. I didn't really care that there was a lack of information as to how the dragonflies can do that; I was more perturbed that the article didn't even go into any detail about how they gathered the data, how they went about exactly finding this out, etc. But yeah, the Discovery article was better, thanks.

      • Re:Hm (Score:2)

        by Piquan ( 49943 )

        except perhaps by reading the article in Nature itself which is not available except by subscription.

        I was under the impression that most college libraries carry Nature. Whether or not most of us would understand the original is another question.

        • Re:Hm (Score:1)

          by Cy Guy ( 56083 )
          I was under the impression that most college libraries carry Nature.

          Yes, I'm sure many do have subscriptions to it. Which doesn't conflict with my statement. My point was that both tthe website, like the print version, requires a subscription to read - that might be your own subscription or someone elses.

          • If you are accessing from within the domain of an educational institution, and your institution subscribes to Nature, you can access the article for free, just like if you went to the library.

            brief article on Nature's site [nature.com]
            At the bottom is an "article" link which takes you to the paper's abstract, and if you have access, you can view the full text via a "full text" link to the left of the abstract.

      • So this is something like a First Post from the scientists?

  • How they manage it still has them puzzled... (Score:5, Interesting)

    by Yarn ( 75 ) on Friday June 06, 2003 @11:30AM (#6132287) Homepage
    I've never seen a puzzled dragonfly. Oh. The scientists.

    I'd assume that the dragonfly merely tries to keep the thing it's hiding from in the same position on *its* retina. It'd be a fairly simple feedback mechanism, if you did it with analogue electronics.
    • I'd assume that the dragonfly merely tries to keep the thing it's hiding from in the same position on *its* retina.

      A nice idea, but it wouldn't work...

      The stalking party can move it's eyes around in it's socket or turn it's head. Either of those would case a change the position on the observing retina, but not on the dragonfly's...

    • No, it's a very different problem. It's only the same if there's only lateral motion; once rotation is involved, it changes drastically.

      Otherwise, it'd be a neat vindication of the Bugblatter Beast.

    • I'd assume that the dragonfly merely tries to keep the thing it's hiding from in the same position on *its* retina.

      But then wouldn't the prey disappear from the drogonfly's vision?

    • Re:How they manage it still has them puzzled... (Score:4, Insightful)

      by barakn ( 641218 ) on Friday June 06, 2003 @02:34PM (#6133904)
      It doesn't seem that simple to me. Imagine a dragonfly flying in a circle around its prey. It yaws appropriately so that it always faces its prey, and so it maintains the prey's image at the same position on its retinae. Instead of appearing to stay at a fixed point to the prey, however, the dragonfly revolves around it a full 360 degrees. This very unstealthy maneuver shows that trying "to keep the thing it's hiding from in the same position on *its* retina" is insufficient.

      In the simplest case, with the prey not moving, all the knowledge the dragonfly needs is the position of the prey. The solution is to fly straight at the prey. It never seems to move from its position on the background but appears larger and larger as it moves in for the kill.

      Cases with the prey moving are more difficult to visualize. You can simplify it by assuming that they are confined to a 2D plane and then drawing their positions on a sheet of paper (or a computer screen). Imagine two diifferent scenarios:

      Case #1. The dragonfly is on a straight line and about half way between the prey and a bush. The dragonfly is superimposed on the bush, from the prey's point of view. The prey is flying perpendicular to this straight line. In order to stay on a straight line between prey and bush, the dragonfly must also move.

      Case #2. Same situation, except that now the dragonfly is practically touching the bush. The prey moves but the dragonfly hardly needs to move at all to appear to remain at the same spot on the bush.

      It should thus be obvious that the distance of the dragonfly to the background object is an important variable. Perhaps it somehow memorizes what object is exactly 180 degrees away from the prey, and then it keeps an eye [plus.com] on both at the same time and flies so as to maintain their positions 180 degrees apart on its retinae (both objects might drift across the retinae, so long as they are exactly opposite each other).

      • Dragonflies tend to predate upon smaller insects, with simple visual systems, so I still think that a kind of pattern maching technique would work. When there's a change in the image of the prey the dragonfly can make a small move in any direction and if it improves it can continue, if it doesn't improve it can revert.
      • The insect visual system is much too primitive to take such variables into account. This is evolutionary economics - because they have such small simple nervous systems that they can't afford to devote too many neurons to any one task. The whole trick is accomplished by visual neurons simply not reacting ( not firing ) to any stimulus that is not changing. So a change in the retinal image will trigger some reaction on the part of the neurons, and bring the attention of the insect to it, but no change wil
    • No that doesn't work. Imagine a line segment AB with A = prey and B = some object. Now imagine that A is constantly moving around. The dragonfly must keep itself on this line despite the prey's motion, and despite the dragonfly's wanting to get closer to the prey. It's not trivial.

      (BTW, I have seen a lot of irrelevant replies to your post. One says something about the swivelling of the prey's eyeballs. Yikes; get a clue.)

    • I think this theory is on the right track! Sure it wouldn't be enough to JUST keep the prey at the same position, but I would guess that anyone who thinks "It must be a lot more complex than that!" is heading down the wrong path. Insects are pretty simple, and simple can be very effective.

      So let's keep things simple. First assume that the prey is simple enough that its eyes don't move relative to its head/body.

      In order to stay in the same spot in the prey's retina, the dragonfly must maintain a positio

  • King Charles II of England did this (Score:4, Interesting)

    by TomGroves ( 622890 ) on Friday June 06, 2003 @11:38AM (#6132362) Homepage
    King Charles 'beheaded' guests who bored or annoyed him by viewing them at such an angle that his blindspot was over their head. Try it for yourself [rmit.edu.au]
    • King Charles 'beheaded' guests who bored or annoyed him by viewing them at such an angle that his blindspot was over their head.

      That's hilarious! I learned about the blind spot in a science class in high school, and from then on I always removed the teacher's head while they were talking. Even into college (although it's been a while since I've practiced; thanks for the memories!).

  • Breaking news: Scientist reinvent the wheel again (Score:5, Interesting)

    by eggstasy ( 458692 ) on Friday June 06, 2003 @11:40AM (#6132382) Journal
    Well duh. Didn't they ever catch flies when they were young? The way to do it is to take two fingers and follow the fly with them, maintaining the distance between your hand and the fly. after a while the fly will think your fingers are part of the background and will easily let you catch it.
    • Well duh. Didn't they ever catch flies when they were young? The way to do it is to take two fingers and follow the fly with them, maintaining the distance between your hand and the fly. after a while the fly will think your fingers are part of the background and will easily let you catch it.

      Yet another example of the universal truth: Everything I need to know, I learned from "The Karate Kid."

      -Isaac

    • The scientists aren't claiming to have discovered the well known fact that stationary objects in a visual field are difficult to see. The surprising discovery is that a primitive insect like a dragon fly is able to take advantage of that knowledge to hide from its prey.

  • Mwa ha! I will be king of the fpers! (Score:4, Funny)

    by dasunt ( 249686 ) on Friday June 06, 2003 @11:56AM (#6132578)

    x <- Moderators, keep staring at this point.
    Everone else can look here -> x

    Now, if my calculations are correct, I should be able to get away with this:
    Imagine a beowulf cluster of F1r5t P05t!

    Mwa ha ha!

    Oh wait, you mean that I'm too big to be a dragonfly?

  • Movie versus Book. (Score:3, Informative)

    by tcak ( 513301 ) on Friday June 06, 2003 @11:57AM (#6132585) Homepage Journal

    Actually the book "The Lost World", which was written by Michael Crichton who wrote "Jurassic Park", shows an opposite behavior of the T-Rex.

    The following lines from the book says:

    Sarah Harding said, "Why did Dodgson just stand there like that? That's not the way to act around predators. You get caught around lions, you make a lot of noise, wave your hands, throw things at them. Try to scare them off. You don't just stand there."
    .....
    "Roxton," Levine said, "believed that tyrannosaurs had a visual system like an amphibian: like a frog. A frog sees motion but doesn't see stillness. But it is quite impossible that a predator such as a tyrannosaur would have a visual system that worked that way. Quite impossible. Because the most common defense of prey animals is to freeze. A deer or something like that, it senses danger, and it freezes. A predator has to be able to see them anyway. And of course a tyrannosaur could."

    • I find Crichton's logic a little shaky here. If "the most common defense of prey animals is to freeze," that suggests that it works a good deal of the time.

      As it stands, this is equivalent to saying that "it's quite impossible that deer are vulnerable to sharp teeth, because the most common offense of predators is sharp teeth."

      (As an editorial nit, how about "In The Lost World, the sequel to Jurrassic Park, Crichton suggests an opposite view.")

      • Yup... Your editorial nit sounds a lot better.

        It's past midnight in Singapore and my mind isn't functioning properly after a hard day at work.
        • Actually, I don't know why I included that last paragraph. Spending too much time on k5, I 'spose. The original was perfectly clear, and it's not like slashdot lets you go back and change things (or else I'd go back and remove my nit-pick.)

          My apologies. Get some sleep. :)

    • It seems that it is now widely accpeted that the t-rex was a scavenger and not a predator. Would seem it would "hunt" by smell more than vision if that was true.
  • I never read the articles and jump straight to the posting as everyone else here. But this seemed interesting so I went and read it. waste of 20 seconds of my time. There are no details! Why not wait until there is something to see before posting these stories?

  • Sounds a little like... (Score:4, Funny)

    by JasonMaggini ( 190142 ) on Friday June 06, 2003 @12:11PM (#6132713)
    The Picard maneuver. Although I doubt dragonflies can punch it up to warp speed yet.
    • Actually from what I can tell it is essentially the same as the Picard Maneuver or at least has the same effect - that the bogey (be it predator or prey) thinks you are stationary when in fact you are moving.

      And it does have military implications [abc.net.au] thought I would think they are limited in that fooling a biological eye is now of little concern - its fooling electronic eyes that is the current challenge, and I can't see how this can be applied to foil them.
  • So that's why my monitor keeps disappearing if I look at it for more than a few mi... oh...
    • So that's why my monitor keeps disappearing if I look at it for more than a few mi... oh...

      Seeing no +1 funnys applied to parent, methinks there has been many-a-"whoosh" over a number of readers' heads....

    • Re:same spot in retina? (Score:3, Informative)

      by zenyu ( 248067 )
      So that's why my monitor keeps disappearing if I look at it for more than a few mi... oh...

      If you take your finger and hold your eyeball in place things will fade to black (so long as you don't move your head and close the other eye.) I don't think this trick would work against us since we can and do move our eyeballs independently of our body. Fireflies only have to pull this trick on the flies they eat...

      The eyemovements we make to be able to sit practically motionless before our monitors is called sa
  • So basically your brain tunes them out, since they're stationary as soon as they come into your field of view, making you tune them out like one might tune out the frames of your glasses, sunglasses, etc.

  • Better Articles (Score:5, Informative)

    by MonkeyBoyo ( 630427 ) on Friday June 06, 2003 @01:31PM (#6133433)
    Boy, that MSNBC [msnbc.com] article was bad. They even mispelled the researcher's name. It is "Akiko Mizutani" not "Aikiko Mizutani".

    Here is some better coverage of the story. discovery [discovery.com], NationalPost [nationalpost.com], and Ananova [ananova.com].

    And here is a nice page [anu.edu.au] from the Insect Vision, Navigation and "Cognition" Laboratory at ANU, but it doesn't cover the dragonfly work.
  • If something was chasing after me, and I had wings, I'd fly away.
    • That's the point. That's why this trick is interesting: it works even if the prey is moving around.

      If it were stationary, the dragonfly would just need to fly straight toward it.

  • A few things guessed... (Score:2, Interesting)

    by Stonan ( 202408 )
    Since there aren't any deatils.

    1. What is are the distances involved?

    2. Best guess, they're using a single lens camera. I believe dragonflies eat flies. If this is so and the fact that flies have compound eyes, does this test really hold true for their natural prey or just for 'human-style' eyes?

    3. I'm not 100% sure myself that dragonflies have compound eyes, but if they do then I would expect that their eyes are accurate enough to see the retina of it's prey (or whatever) and keep itself in the same pos
    • Dragonflies catch many insects, not just flies. You're right that most bugs have compound eyes, though, so it is an interesting question.

      A point, though... they wouldn't have to be thinking "human-style" eyes. It's probably more of a defensive measure than an offensive measure, though... as birds have normal eyes, not compound eyes. Their eyes are also, largely, on the sides of their heads, and wouldn't give them great depth perception.

      Still... a very interesting idea.

  • From reading the Nature article (Score:5, Informative)

    by Dachannien ( 617929 ) on Friday June 06, 2003 @03:53PM (#6134616)
    Generally speaking, the dragonfly moves in such a way that if you draw a line from the dragonfly to the prey at each increment of some time step, the lines will (nearly, because it's not perfect) cross at one point. Thus, to the prey, it appears that the dragonfly is a stationary object located at the point where the lines cross.

    It relies on a lack of depth perception, obviously. As a guess, perhaps the dragonfly is able to accomplish this by using the same visual cues it evokes in its prey - if the dragonfly moves in the right way, then its prey will appear to be a stationary object (from the dragonfly's perspective) as well.

    However, this doesn't account for situations where the dragonfly emulates an object that is behind it (i.e., the lines cross at a point on the far side of the dragonfly) or an object at a large distance (where the dragonfly directly shadows the prey, copying its every move).

    If you are still confused, think of it this way: You're playing your favorite first-person shooter, and you want to hide behind a tree/pillar/rock so that an approaching target can't see you. You can move around the tree so that it always forms an intervening object. If you draw a line between yourself and your target at each moment in time, they all intersect at the tree. If your target happened to have really crappy eyesight (compound eyes, perhaps) then you could just remove the tree, and at every moment in time they'd see you there along the same line of sight where the tree would have been, so the target perceives you as being located where the tree would have been and moving along as if you were a part of the landscape. (The advantage, though, is that you can move around and close in on your prey, while your prey remains unaware of the soon-to-occur frag.)
    • In case you want to try to access the article:

      Subscription may be required to read the full article [nature.com]

    • Dragonflies maybe have near 360 degree vision, considering their large bulbous compound eyes, which might account for the rear-view capabilities?

      In any case, the real question is how we as humans intend to take advantage of this knowledge when dealing with similarly equipped, ie steroscopic vision, opponents. It's not as if we can actually move that fast, at least not at close range. Maybe that's it, maybe it's something we could use for our stealth program.... nah nobody uses 'video' for aircraft detectio
    • As a guess, perhaps the dragonfly is able to accomplish this by using the same visual cues it evokes in its prey - if the dragonfly moves in the right way, then its prey will appear to be a stationary object (from the dragonfly's perspective) as well.

      This wouldn't work - in the Discovery article, they said that this study was conducted with pairs of male dragonflies jousting for territory. Obviously then this technique works on the dragonflies own visual system as well as that of insects it might prey u

  • I'm sure i'm probably too late... someone's probably already applied for a patent on dragonfly flight.... which means I'd be infringing if i walk a certain way, i guess....

  • Background tracking error reduction (Score:3, Interesting)

    by Anonymous Coward on Saturday June 07, 2003 @09:10PM (#6140667)
    I've seen something like this before. Once I was outside cleaning up after a party. I went to pick up a vase of flowers and I noticed a few tiny fruit flies(?) that were hovering near the flowers. The funny thing was that when I picked up the flowers, these flies would maintain the exact same relative position to the flowers. Even if I rotated the vase around its axis.

    It was like taking the flies for a walk on an imaginary but invisible leash.

    I guess that the flies had an instinct that to remain still, they must reduce the error in *their* retina between the current background image and the stored background image. I am guessing that dragonflies have evolved to do the same thing but with a greater degree of freedom. i.e. a chosen target rather than the whole background.
  • The article suggests that armies could learn something useful about camouflage from dragonflies. What are they suggesting, that soldiers flit about the battlefield to keep in the same position in the enemies' retinas all the time - how would they manage this? Jet packs? Pogosticks? Even a helicopter which has maneuverability analogous to a dragonfly could never do this. The scale for humans is just all wrong - a dragonfly only has to dart short distances, because the relative scale is quite small, but a
  • ...is (probably) that the prey has fixed eyes, rather than mobile ones. This way, the dragonfly only needs to take the prey's body (or head, if it is not fixed, too) position into account. Otherwise the problem becomes much, much more difficult, as eye movement is generally quite a bit faster than head or body movement.

    There is also a question of spatial resolution of the prey's visual system to consider, as well. A more coarse-grained vision (i.e. less photoreceptors/mm^2) would be easier to fool than

Slashdot Top Deals

Too much is not enough.

Close