 |
 | Home | Contact Us | FAQ | Search & Site Map | Link to Us |
 | Sign In | Join | Other 45 Sites in Network |
Medical Forum / General / Vision / May 2009Eyes Rod/Cones vs Digicam Sensels
Hello guys, How do the eyes rods and cones compare to a digicam sensor in dimensions? A typical digicam has a sensel (sensing element or commonly known as pixel) size of 0.005mm or 5 micron, how about the rods and cones? What are their sizes? Are they round or square and what is the separation between them? Digicam sensor pixels have no separation, our eyes rods/cones have separations, why do we see whole and not lines. Also I heard our eyes have resolution of 1 arcminute, how is it calculated from the rods/cones sizes and separations? Many thanks! Ann The rods and cones don't work independently, like camera elements. In the retina, rods and cones are connected by "horizontal" cells that gang together several sensing cells to create "receptive fields". There aren't nearly enough neurons in the optic nerve to transmit one-to-one raw "pixel" data. What goes to the brain is a derivative, a signal that a group of receptors, working together, detected a light-dark border with a certain orientation (and/or moving in a certain direction.) I think the effective "data compression" is about 1000:1. Cone cell diameters range from .5 to 4 microns. But receptive fields can be as large as 1500 microns (1.5 mm!) Receptive fields overlap, and each cone/rod can participate in many receptive fields. The derived data is collected by "ganglion" cells that most directly connect to the optic nerve. -MT > Hello guys, > [quoted text clipped - 14 lines] > > Ann > The rods and cones don't work independently, like camera elements. In the > retina, rods and cones are connected by "horizontal" cells that gang [quoted text clipped - 35 lines] > > - Show quoted text - Hi Dr. Mike, Do you agree with the following web site: http://www.clarkvision.com/imagedetail/eye-resolution.html LIke how our eyes have dynamic range of 1:1,000,000 or how it is equal to more than 576 megapixels in quality? Some folks believe the site is in error and the eyes equivalent megapixel is just 1 Megapixel and dynamic range low. What is your say about it? Thanks. A > Do you agree with the following web site: > [quoted text clipped - 3 lines] > or how it is equal to more than 576 megapixels in > quality? Skimmed through it and it seems pretty well-founded. > Some folks believe the site is in error and the > eyes equivalent megapixel is just 1 Megapixel > and dynamic range low. What is your say about it? > Thanks. The dynamic range, well, you can't really dispute that. Full-sun or more at one end, moonlight at the other, that's a lot of f-stops. The megapixels thing makes at least one wrong assumption. There's a huge difference in receptor density between the center and the periphery. If we say the resolution approximates a 5 MP camera in the center, then at the edges, it's more like a motion detector. See http://tinyurl.com/conedensity -MT > > Do you agree with the following web site: > [quoted text clipped - 22 lines] > > -MT So how do you think they can create artificial eyes? If say the cones and rods or retinas are damaged, then they have to design and put ccds where the output can be connected to the optic nerve. And if the optic nerve is damaged, they can directly connect the output of the CCD to the visual cortex itself. Know what is the latest in artificial eye? Our microprocessor processing center like INTEL already have the capability to create such micro stuctures in the scale of artificial rods and cones. When do you think signicant breakthrough can occur? And what's stopping it? A > > > Do you agree with the following web site: > [quoted text clipped - 34 lines] > scale of artificial rods and cones. When do you think > signicant breakthrough can occur? And what's stopping it? The tricky bit is hooking up the sensors to the visual cortex. In the eye, the rods and cones do not send their output directly to the cortex, the other layers of the retina do quite a bit of processing before sending the signal to the cortex. Dr Judy The tricky bit is hooking up the sensors to the visual cortex. In the eye, the rods and cones do not send their output directly to the cortex, the other layers of the retina do quite a bit of processing before sending the signal to the cortex. Dr Judy Wonder where the natural blind spot is computed out. Don W. > Wonder where the natural blind spot is computed out. Good question. Acquired blind spots are also concealed. It might be a natural consequence of retinal pre-processing. The entire visual field fades to gray if you immobilize the image. In both cases, no edges, no motion, no color equals no awareness. -MT > Good question. Acquired blind spots are also concealed. Well, scotomas "acquired" during macular degeneration? > It might be a natural consequence of retinal pre-processing. The entire > visual field fades to gray if you immobilize the image. In both cases, no > edges, no motion, no color equals no awareness. What is truly amazing is the patching in of a crazy quilt design into the natural blind spot area. Don W. > Well, scotomas "acquired" during macular degeneration? That's my understanding. It doesn't fade to black, just a grayish lack of perception, an annoying void. >> It might be a natural consequence of retinal pre-processing. The entire >> visual field fades to gray if you immobilize the image. In both cases, no >> edges, no motion, no color equals no awareness. > > What is truly amazing is the patching in of a crazy quilt design into the > natural blind spot area. I'm lost - I'm not familiar with that. -MT >> What is truly amazing is the patching in of a crazy quilt design into the >> natural blind spot area. > > I'm lost - I'm not familiar with that. > > -MT Mike, All I was "appreciating" was the way in which the blind spot is covered in (up?), in normal everyday living. Easy (easier?) to understand when a plain vanilla surface is extrapolated into the natural blind spot. But.... when a crazy quilt (all kinds of patches, squares of colors) is patched in ... (i.e., with no holes seen) is truly amazing. Don W. Don W <dwilgus@prodigy.net> wrote in part: >>> What is truly amazing is the patching in of a crazy quilt design into the >>> natural blind spot area. [quoted text clipped - 8 lines] > crazy quilt (all kinds of patches, squares of colors) is patched in ... > (i.e., with no holes seen) is truly amazing. I think you missed Mike's point about movement being _required_ for vision. No change on the cone, no signal. The eyes almost always are in constant motion to see and incidentally fill in the [small] blind spot. When the eyes are "fixed", then they become very sensitive to motion anywhere in the visual field. Very useful. -- Robert > I think you missed Mike's point about movement being _required_ > for vision. No change on the cone, no signal. The eyes almost [quoted text clipped - 5 lines] > > -- Robert Robert, That "natural blind spot", 5 to 7 degrees, anything but "small" (to me). I remember reading about experiments with a silvered spot on a contact lens to help fix the received visual field. The field was said to fade. But with that (I am extrapolating your comment) when a pixel moves anywhere in the fixed visual field, it stands out? So rods are also sensitive to that change effect also, I presume. But I'm just talking about the computational fill in properties of the natural optic disc blind spot. (Without help from the other eye). Under "normal" viewing conditions. Don W. > That "natural blind spot", 5 to 7 degrees, anything but "small" (to me). If that was 7 degrees out of 160, it would be "small". But your visual field is a two-dimensional area, not a horizontal line. It isn't 7/160. It's pi x (3.5 ^2) out of pi x (80^2). That's ~40 square degrees compared to ~20000. Or as we say in Alabama, diddly-squat. > I remember reading about experiments with a silvered spot on a contact > lens to help fix the received visual field. > The field was said to fade. Other methods for immobilizing the image all say the same. > But with that (I am extrapolating your comment) when a pixel moves > anywhere in the fixed visual field, it stands out? > So rods are also sensitive to that change effect also, I presume. Rods can only do one thing - fire in the presence of light. Each time they fire, they exhaust their ability to fire again, a "refractory" period. After they recharge, they fire again if the light is still there. Otherwise they don't. One rod cannot distinguish direction or movement. All it can do is fire in the presence of light and stop firing when light is gone. Anything more than that must be detected by horizontal connections. There are independent little networks of rods or cones, with horizontal and amacrine cells comparing one part of the group with another. This forms a "receptive field" with the ability to detect lines, orientation, and direction of movement. > But I'm just talking about the computational fill in properties of the > natural optic disc blind spot. Most of us can't tell there's anything missing. -MT >> But I'm just talking about the computational fill in properties of the >> natural optic disc blind spot. > > Most of us can't tell there's anything missing. > > -MT I think I know of one situation ("where we can't tell there's anything missing ") that has resulted in an accident. Driver comes to a stop in an intersection and is going to make a left turn. Driver looks _once_ to the left, not turning the head completely, so that monovision only is to the left. (Nose blocks completely bioptic vision). Traffic (one car) is coming from the left and is visually trapped in the blind spot. Of course, we can't tell if something is missing since blind spot patches in the optic nerve blind spot. Driver proceeds, collision occurs. My theory as to how that accident occurred to a very careful driver. Don W. > > >> But I'm just talking about the computational fill in properties of the [quoted text clipped - 19 lines] > > My theory as to how that accident occurred to a very careful driver. Looking to the left establishes whether a car is there (warning to brain!). For a normal eye, even if you aren't looking to the right or left, the peripheral vision will detect any motion. That's my counter-theory, although I'm not too confident about it.  SignatureDan Abel Petaluma, California USA dabel@sonic.net Don W <dwilgus@prodigy.net> wrote in part: > That "natural blind spot", 5 to 7 degrees, anything but "small" > (to me). Of course your judgement is up to you. But 5-7' on a visual field of ~120' horiz * ~100' vert is only 0.3% . > I remember reading about experiments with a silvered spot on a > contact lens to help fix the received visual field. The field > was said to fade. But with that (I am extrapolating your > comment) when a pixel moves anywhere in the fixed visual field, > it stands out? That sounds therapeutic for some sort of ocular pathology. > So rods are also sensitive to that change effect > also, I presume. I would presume so as well. rods are very motion sensitive (as well as blue-sensitive). > But I'm just talking about the computational fill in properties > of the natural optic disc blind spot. (Without help from the > other eye). Under "normal" viewing conditions. Movement of 6' isn't much. -- Robert > Also I heard our eyes have resolution of 1 arcminute, > how is it calculated from the rods/cones sizes and > separations? I hit Send too early... It isn't calculated, it's measured, and I think you meant one arc-SECOND. The best performance comes from a test called "vernier acuity" - detecting the offset of a line segment, like __________----____________ Instead of measuring, if you know the diameter of a cone, you could calculate a theoretical value with trigonometry. If you assume one cone is 1 micron, then you'd imagine a right triangle with the other leg hmmm 2 cm, extending from the retina to the crystalline lens. =20000 microns. Then, for the angle at the lens, the opposite side is 1 and the hypotenuse ~20000. Opposite over hypotenuse is sine. The arcsin of 1/20000 is.... 0.002864789 degrees..... 10 arc seconds? I'm making assumptions and I'm not checking my work but that sounds right, thereabouts. The Titmus Stereo card used in many offices tests down to 20 arc seconds so I know wer'e in the right ballpark. Receptive fields can cloud this issue but they work together, via comparisons make at the cortex. Your brain uses the fact that _this_ group of fields got stimulated while _that_ group did not, to regain the resolution you'd expect from pixel-by-pixel interpretation. The receptive fields also participate to give sensitivity to low levels of light, so that the retina is capable of detecting just a few photons as "light." -MT |
Reply to this Message |
 Sign In
 Join
 My Latest Posts My Monitored Threads My Blog My Photo Gallery My Profile My Homepage Start New Thread Enable EMail Alerts Rate this Thread
|
©2009 Advenet LLC Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.