Some Interesting Things About Flies..

Images of undisturbed subjects in their natural environment. All subject types.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

anvancy
Posts: 387
Joined: Sat Dec 05, 2009 11:57 pm
Location: India
Contact:

Some Interesting Things About Flies..

Post by anvancy »

Image

Ever wondered why a fly flies or makes circles around food for en-number of times?
Sure it is after the food.But the fascinating part is,they have taste receptors on their feet and on the tips of the wings.So when they are flying and stand/sit on something they are actually tasting that particular subject. Tiny hairs on the feet of the fly relay taste signals to its basic brain telling the various flavours it can savor on.

Image

Flies do blow bubbles to clean their mouthparts with this fluid, more or less to flush them, and after they are ready they ingest the fluid again.
Flies can use bubbling for cooling of overheated body ? thermoregulation.

Image

Insect eyes like those of robberflies,houseflies,wasps etc all have a peculiar pattern.But wary in the terms of image processing.

Take the case of this housefly.

The main eyes are made up of huge number of small eyes.So each dot you see on the eye structure is one eye.Its like hundreds of miniature cameras all synced up and transmitting images constantly.

The interesting part is,due to this structure,flies see a combination of a mosaic.So they dont see a clear image,but a mosaic of images.Upon this,due to their body structure,the brain is quite small considering the eye structure.Due to this,the fly sees us as stop animation subjects.It doesnt see as 30fps or so.It sees frozen images moving in between.Like CCTV footage.

Pretty awesome indeed.

If you have some more facts,things do share..

Anvancy
www.anvancy.com

Raynox 150|Raynox 250|Raynox MSN 202|Canon MPE 65mm|Canon 100mm.|Wemacro Rail

LordV
Posts: 1571
Joined: Thu Nov 22, 2007 10:28 am
Location: UK

Post by LordV »

Lovely series. Quite a debate on fly bubble blowing activity here http://www.diptera.info/forum/viewthrea ... ad_id=1016
Brian v.
www.flickr.com/photos/lordv
canon20D,350D,40D,5Dmk2, sigma 105mm EX, Tamron 90mm, canon MPE-65

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Some Interesting Things About Flies..

Post by rjlittlefield »

anvancy wrote:The main eyes are made up of huge number of small eyes.So each dot you see on the eye structure is one eye.Its like hundreds of miniature cameras all synced up and transmitting images constantly.

The interesting part is,due to this structure,flies see a combination of a mosaic.So they dont see a clear image,but a mosaic of images.
Sorry to object, but this description is simply not correct.

Each facet of the fly's eye (called an ommatidium) is like a single pixel in a digital image. It provides a signal that represents the average brightness of some small region around the fly. These regions are basically cones that extend perpendicular to the overall surface of the eye and overlap to a considerable extent in the far field. Thus what the fly sees is not a mosaic of images, but instead a single image that is rather blurred.

Behind the segmented surface of the eye, the photoreceptors are interconnected in an essentially 2D structure that is very similar to the human retina.

In both cases the "hemispherical world" seen by the eye is optically projected so that it maps onto a curved 2D array of photoreceptors. The primary difference is that in the human eye, the projection is done by a single large wide angle lens, while in the fly eye, the projection is done by an array of small narrow angle lenses.

While the structure appears very much different, there are great similarities behind the lens(es).

--Rik

sonyalpha
Posts: 915
Joined: Sat Apr 24, 2010 1:41 am
Location: Middle England
Contact:

Post by sonyalpha »

Just fascinating....................a great discussion based upon some wonderful photography...................I can add nothing but admiration:

sonyalpha
Retired but not old in spirit:

Fairly new to photography........keen to learn:

anvancy
Posts: 387
Joined: Sat Dec 05, 2009 11:57 pm
Location: India
Contact:

Re: Some Interesting Things About Flies..

Post by anvancy »

rjlittlefield wrote: Sorry to object, but this description is simply not correct.

Each facet of the fly's eye (called an ommatidium) is like a single pixel in a digital image. It provides a signal that represents the average brightness of some small region around the fly. These regions are basically cones that extend perpendicular to the overall surface of the eye and overlap to a considerable extent in the far field. Thus what the fly sees is not a mosaic of images, but instead a single image that is rather blurred.

Behind the segmented surface of the eye, the photoreceptors are interconnected in an essentially 2D structure that is very similar to the human retina.

In both cases the "hemispherical world" seen by the eye is optically projected so that it maps onto a curved 2D array of photoreceptors. The primary difference is that in the human eye, the projection is done by a single large wide angle lens, while in the fly eye, the projection is done by an array of small narrow angle lenses.

While the structure appears very much different, there are great similarities behind the lens(es).

--Rik
Thank you for correcting the knowledge.This is what I like.Factual things.It seems then Animal Planet told me wrong.

Question to you Rik,Do flies see whole time 2D or they have a processor to do 3D imaging?Do flies see flat images without any depth?

And thanks Brian and Sonyalpha!

Anvancy
www.anvancy.com

Raynox 150|Raynox 250|Raynox MSN 202|Canon MPE 65mm|Canon 100mm.|Wemacro Rail

Retro
Posts: 57
Joined: Wed Mar 31, 2010 5:15 pm
Location: Scarborough, Ontario

Post by Retro »

Fantastic photos Anvancy, :D as all your contributions thus far have been. I very much appreciate your insights as well.

rjlittlefield wrote:
Behind the segmented surface of the eye, the photoreceptors are interconnected in an essentially 2D structure that is very similar to the human retina.
That supports what I thought (and wrote) about when looking at rumbert's bee photo; supported by Tesselator's opinion and reference links. For me this is a very radical paradigm shift.

Keep shooting,
Jim
Last edited by Retro on Sat Aug 21, 2010 6:07 pm, edited 1 time in total.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: Some Interesting Things About Flies..

Post by rjlittlefield »

anvancy wrote:Question to you Rik,Do flies see whole time 2D or they have a processor to do 3D imaging?Do flies see flat images without any depth?
It is a subtle question. At some point we need to decide what "3D imaging" means.

Because of media hype, we are conditioned to think that "3D" is synonymous with "binocular stereo". In this process, 3D structure is inferred by analyzing differences between two images taken from different viewpoints.

I am not aware of any evidence that flies use binocular stereo.

But from a functional standpoint, binocular stereo is certainly not a requirement for 3D perception. For example, a few percent of humans do not have functioning stereo vision either, and those individuals are not much impeded by that lack. This is because there are many other ways to infer depth.

Other ways to infer depth include shadowing and brightness gradients ("shape from shading"), occlusion (front/back overlap), and optical flow. Occlusion is often talked about because it seems so obvious and clearcut. We older humans are quite sure that when one object disappears at the edge of another, the object that disappeared is simply farther away and went behind the other. However, this understanding requires quite sophisticated processing -- even humans have trouble interpreting "disappeared behind" when we are young enough. Optical flow is both simpler and more reliable. Assuming the world is static, then when you move your head, things that are closer appear to move faster (more degrees per second). This is optical flow, and it has the wonderful feature that you don't even need "objects" to make it work. All that's needed is some ability to correlate local textures.

I am not aware of any evidence that flies use either shape from shading or occlusion. Maybe they do, maybe they don't. Even in humans, it's not easy to design experiments to pick apart the various methods.

Flies certainly do use optical flow, however, and they are very good at it. If you do a Google search on fly optical flow computation, you will find lots of articles with detailed discussions. Interestingly, flies do use binocular comparison to help analyze the optical flow information (see HERE, but this is not the same as binocular stereo, which also works on static images.

Getting back to your questions, I would say that flies definitely see 3D when they are moving, but perhaps not when they are at rest!

--Rik

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Retro wrote:That supports what I thought (and wrote) about when looking at rumbert's bee photo; supported by Tesselator's opinion and reference links. For me this is a very radical paradigm shift.
I missed that particular conversation. I do, however, vividly recall a similar sensation when I first encountered a detailed description of a fly's eye, something like 5 years ago. Prior to that, compound eyes had always been troubling to me because I was stuck in the "mosaic of images" model. When I got popped over to the"mosaic of lenses" model, life suddenly got a lot simpler.

--Rik

anvancy
Posts: 387
Joined: Sat Dec 05, 2009 11:57 pm
Location: India
Contact:

Re: Some Interesting Things About Flies..

Post by anvancy »

rjlittlefield wrote:
anvancy wrote:Question to you Rik,Do flies see whole time 2D or they have a processor to do 3D imaging?Do flies see flat images without any depth?
It is a subtle question. At some point we need to decide what "3D imaging" means.

Because of media hype, we are conditioned to think that "3D" is synonymous with "binocular stereo". In this process, 3D structure is inferred by analyzing differences between two images taken from different viewpoints.

I am not aware of any evidence that flies use binocular stereo.

But from a functional standpoint, binocular stereo is certainly not a requirement for 3D perception. For example, a few percent of humans do not have functioning stereo vision either, and those individuals are not much impeded by that lack. This is because there are many other ways to infer depth.

Other ways to infer depth include shadowing and brightness gradients ("shape from shading"), occlusion (front/back overlap), and optical flow. Occlusion is often talked about because it seems so obvious and clearcut. We older humans are quite sure that when one object disappears at the edge of another, the object that disappeared is simply farther away and went behind the other. However, this understanding requires quite sophisticated processing -- even humans have trouble interpreting "disappeared behind" when we are young enough. Optical flow is both simpler and more reliable. Assuming the world is static, then when you move your head, things that are closer appear to move faster (more degrees per second). This is optical flow, and it has the wonderful feature that you don't even need "objects" to make it work. All that's needed is some ability to correlate local textures.

I am not aware of any evidence that flies use either shape from shading or occlusion. Maybe they do, maybe they don't. Even in humans, it's not easy to design experiments to pick apart the various methods.

Flies certainly do use optical flow, however, and they are very good at it. If you do a Google search on fly optical flow computation, you will find lots of articles with detailed discussions. Interestingly, flies do use binocular comparison to help analyze the optical flow information (see HERE, but this is not the same as binocular stereo, which also works on static images.

Getting back to your questions, I would say that flies definitely see 3D when they are moving, but perhaps not when they are at rest!

--Rik
Interesting.

to my knowledge,eyes occupy a large part of a fly's body from the front.The brain is usually small correct?If this is small then the fly may not have the necessary "processing power" to add depth each and every second.

And thanks for updating my knowledge Rik.Never knew there were mainly three types of 3d detection systems.

That day I read an article about 3D films and its profit making business.They said that in,the average crowd,60% of the viewers cannot interpret the sudden 3D movement since they cannot process it quickly.Since they cannot process it,they are unable to enjoy the real 3D experience.Just an OT.

Will surely check that link out.

And thanks Jim.Always trying to improvise.

Anvancy
www.anvancy.com

Raynox 150|Raynox 250|Raynox MSN 202|Canon MPE 65mm|Canon 100mm.|Wemacro Rail

gmazza
Posts: 566
Joined: Sat Dec 12, 2009 4:03 pm
Location: Rio Grande do Sul; Brazil; 29°S 51°W

Post by gmazza »

Very interesting info about the vision of compound eyes, I wonder about the function of the tree omantidia in the middle of head, between the compound eyes, what these add to overall perception ? Why this strategic position ?
Gustavo Mazzarollo

Portfolio

http://www.gmazza.com

cozdas
Posts: 68
Joined: Fri Jun 11, 2010 9:42 pm

Post by cozdas »

After taking the broken glass image, I decided to learn more about the compound eye structures and did some web-reading here and there.

My previous knowledge was that each small lens (ommatidium) is associated with only one photo receptor, thus the resolution of an insect's eye is directly proportional to the number of ommitidia. What I learned while reading is slightly different from that; of course the nature is more complicated.

There are different "designs" (as one pro-creationist web site that I read calls - not that I agree with their conclusions but the info there looks valid). Wiki says there are two main groups: Apposition eyes and Superposition eyes. Looks like especially in the superposition eyes the lens-photoreceptor relation is pretty complex (not one-to-one). With "Neural Superposition" eyes lens-photoreceptor relation is one-to-many, where each lens indeed creates a very low res version image which are then combined to generate the whole image. But still the resolution is not as simple as (num lenses) x (num photoreceptrs per lens) because photoreceptor cell nerves from different ommatidia are connected additively to increase the light sensitivity.

Image
... In the other kind of apposition eye, found in the Strepsiptera, lenses are not fused to one another, and each forms an entire image; these images are combined in the brain. This is called the schizochroal compound eye or the neural superposition eye. Because images are combined additively, this arrangement allows vision under lower light levels.
So for these kinds of compound eyes the overall optical system seems to be like that:
- each ommatidium corresponds to a partial cone of the entire field of view but those cones are not mutually exclusive meaning those cones from different ommatidia are overlapping.
- photoreceptors in each ommatidium corresponds to (probably mutually exclusive) fraction of the cone of that ommitidium.
- Some some receptors of an ommitidium shares the same view direction with some other receptors from some (but not all) other ommitidium. In other words a particular view direction is represented by some photoreceptors in few ommitidia.

it's a pretty similar to 4D light field structure that I'm familiar with in computer graphics, but unlike the light field sensors, compound eyes are not flat but curved thus low-res sub images does not cover the same FOV.

Image

The bottom line is I learned that my "one lens = one photoreceptor" knowledge was valid only for some sub-set of the compound eyes. While reading that I also remembered this thread and re-read Rik's explanation. I guess he was describing one of the other superpostion eye structures.

Image
Image
Image

I thought I should share what I learned, it's an amazing subject. Hopefully I grokked it correctly :) Please feel free to correct me if I misunderstood some stuff.

Harold Gough
Posts: 5786
Joined: Sun Mar 09, 2008 2:17 am
Location: Reading, Berkshire, England

Re: Some Interesting Things About Flies..

Post by Harold Gough »

anvancy wrote: They said that in,the average crowd,60% of the viewers cannot interpret the sudden 3D movement since they cannot process it quickly.
I suspect something similar might affect some drivers, who may not all be aware. I know for sure than my son has excellent vison in both eyes but reports rarely seeing a 3-D effect, even with static stereo pairs. It certainly makes him hesitant when (practice) driving.

Harold
My images are a medium for sharing some of my experiences: they are not me.

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

Whilst I suppose scientists are very clever, the basic fact is they are not flies so do not know how a flies brain sees the world and can only speculate. We have two eyes but our brain interprets them as a single image since basically you see with your brain because it interprets the images it receives from your eyes. We know this is true because we can discuss it with each other and confirm we all are seeing in a similar manner.

I don't know of any flies or other creatures with multifaceted eyes that have discussed what their brain sees with scientists, so scientists really can only speculate as to what they see. If our brains can combine the different images from two eyes into a single image why should not a flies brain be able to combine all the images it receives similarly? The truth is we shall probably never know the answer and can only speculate.

I am looking with two separate eyes at a computer screen composed of millions of individual pixels analogous to signals from a multifaceted eye but my brain is combining them into one comprehensible picture, why should not a flies?

Science relies on theories and hypotheses that seemingly fit the facts until disproved by some later theory, unless it can be categorically proved at the time.

DaveW :)

cozdas
Posts: 68
Joined: Fri Jun 11, 2010 9:42 pm

Post by cozdas »

We have two eyes but our brain interprets them as a single image since basically you see with your brain because it interprets the images it receives from your eyes.
I don't think that we can call the combined information from two eyes as "single image". Image suggests 2D flat picture, it's not what our brains perceive, it's more elaborate.

Eyes and other senses are just sensory inputs which feed the brain with data and brain in return creates an internal "guessed model" of the environment which is constantly updated with the streaming new data. This model is built using lots of heuristics, interpolations and learned/hard-coded information, so there is no one-to-one seen-perceived correspondence. You measure the light spectrum only with 3 (4 if you count rod cells) types of cells but you perceive a continuous color scale, R, G, B and lightness measurements are happening in different locations in your eye but you perceive smooth colors not polka-dot patterns. Your eyes have considerably low number of blue cone cells than green ones but you don't realize that your blue channel is recording in lower resolution. Or you are not aware of your blind-spot. Why? because what you "perceive" is not the signal that are coming from your eyes, what you perceive is the internal "model" of the surrounding in your brain which is built with assumptions, interpolations etc. This is why it's possible to see weird things as in "optical illusions", those happen when the assumptions to build that model fail.

You said "you see with your brain"; I totally agree. Once you start thinking the visual perception as an model of the world instead of a photographic snapshot, then different methods and sensors to acquire the data to build that model becomes less important in terms of perception. I don't think that insects see the surroundings as image tiles even if they have many photo-receptors per ommitidium. I believe that each an every animal has similar "model" of their surroundings in their brains which hides the shortcomings of the sensory input systems that they use to built that model. So I don't think that animals/insects etc "see" (or more correctly "perceive") the surroundings much different than us do. I think this applies even to bats who have switched importance of visual vs auditory senses.
We know this is true because we can discuss it with each other and confirm we all are seeing in a similar manner.
Not entirely; there are lots of research done on the physiology starting from cell level to up. We know how the RGB channels are combined with each other before sending the signal to the brain, this for example explains why there is bluish-green or greenish-blue but not "reddish-green" ;o)

DaveW
Posts: 1702
Joined: Fri Aug 04, 2006 4:29 am
Location: Nottingham, UK

Post by DaveW »

I was using "single image" in the sense that we do not see a separate image from each of our eyes side by side, but rather analogous to photostacking in that out brains "stacking software" combines them to produce a more 3D image. Therefore I see no reason why an insects brain does not perform a similar combination of separate images to form a similar 3D version of what is around it.

I cannot believe for instance a dragonfly can be so good at catching insects in flight if it is only seeing a rudimentary fuzzy image?

DaveW

Post Reply Previous topicNext topic