How photons create an image via the brain

[boffinwannabe]

im trying to understand how photons allow the brain to create an image. I understand that light is made up of photons. These photons are emitted from an energy source and reflect off an object and hit the retina which produces electrical signals that the brain creates an image from. What i don't understand is how the photons transfer the information of say a tree to the eye. What changes from just before the photon hits the tree to just after? what allows the photon to transfer the information of the tree to the eye/brain?

 

[zhermes]

An individual photon doesn't say "tree," it says what color it is (based on the frequency, which is determined by the surface from which it was emitted/reflected) and where it came from (based on the trajectory determined from which part of the retina the photon hit).

Before the photons (many many many many of them) hit the tree, they're all randomly oriented and all colors (frequencies). After they hit the tree, the ones that will later hit your eye are particular colors coming from particular directions.

 

[ebits21]

Lets say you have a green leaf. Light of various wavelengths are emitted from the sun and hit the leaf. Most are absorbed, but photons with wavelengths that are "green" are reflected off of the leaf and enter your eye.

Your retina then has various cells called rods and cones that have proteins called opsin associated with them. The photon interacts with opsin, which has a chain of effects that leads to an electrical signal that is processed in your visual cortex at the back of your brain. Rods and cones have a spatial layout on your retina that the brain "knows", and can therefore reconstruct an image.

So in a nutshell, the photon has a wavelength that your brain interprets as color. Further, the photons spatial position tells your brain the shape of the object you're looking at. Shape + color = the image of the object.

 

[boffinwannabe]

now when you say the wavelength of the photon, as determined by what? We have a stream of photons and the wavelength is determined by the frequency between each photon? how do we get a situation of various wavelengths? i understand photons have duality its just hard for me to understand what amounts to full spectrum light. Is it a wave like a sea wave or just the distance between two photons? And a stream of photons will have varying distances between them?
anyway thanks for the explanations of above. I can understand how we get a colour from the wavelength, but how we can reconstruct an image from a series of photons is currently beyond me. :)

 

[ebits21]

You can think of the wavelength as the length between peaks on the wave (or troughs, or the period of a trigonometry function is you've covered that). Frequency is how fast those peaks (or whatever) are cycled through. Because the speed of light is constant the two depend on each other, so in the case of light you can think of wavelength and frequency as describing basically the same thing. In other words, the frequency will always be the same for a given wavelength. (This isn't true for all non-light waves where the speed is NOT a constant).

White light, like from the mid-day sun, is actually a collection of photons that have many different frequencies/wavelengths. Think of a prism. If you pass white light through it you get all the colors of the rainbow. Your brain just interprets this varied collection of photons as white.

The white light (all colors) from the sun hits the leaf, but only photons with a particular wavelength/frequency (green) "bounce" off in a direction headed towards your eye.

Your retina is like a grid of receptors, so the place that the photon ends up on the "grid" in the back of your eye depends on where it came from in the "real world". The image on your retina is actually upside down and inverted.

You can also use this analogy. Your computer monitor is a grid of red, green, and blue pixels all laid out on the screen (just like your retina is a grid of red, green, and blue receptors). To make an image, the monitor lights up certain pixels at certain locations in the grid. Similarly, receptors on the retina register particular colors at certain spots on the grid in the back of your eye.

 

[ebits21]

Oh, and also. Each photon behaves as both a particle AND a wave. So the frequency/wavelength doesn't have to be between two photons in a stream. Even a single photon will have a frequency/wavelength just by itself.

It should be a little confusing because water and sound waves that you're familiar with do not behave like that. A single water molecule will not create a wave. Things in the quantum World can be very confusing.

Edit: These videos may be of some help if you're confused still.
http://www.khanacademy.org/video/introduction-to-waves?playlist=Physics"
http://www.khanacademy.org/video/amplitude--period--frequency-and-wavelength-of-periodic-waves?playlist=Physics"

 

[Andy Resnick]

im trying to understand how photons allow the brain to create an image. I understand that light is made up of photons. These photons are emitted from an energy source and reflect off an object and hit the retina which produces electrical signals that the brain creates an image from. What i don't understand is how the photons transfer the information of say a tree to the eye. What changes from just before the photon hits the tree to just after? what allows the photon to transfer the information of the tree to the eye/brain?

The photoreceptors in your eye are the first layer in a highly complex computational scheme (the visual cortex). Within your retina, there are approximately 5 processing steps that occur (local averaging, edge detection, etc) and this information is passed to the lateral geniculate nucleus which decodes motion and depth perception (among other things), and then on to the visual cortex. The elemental stimulus of the visual cortex (what the neurons in V1 respond to) seems to be oriented line segments.

A good book for this is "Basic Vision", by Snowden, Thompson, and Troscianko.

 

[Neandethal00]

The photoreceptors in your eye are the first layer in a highly complex computational scheme (the visual cortex). Within your retina, there are approximately 5 processing steps that occur (local averaging, edge detection, etc) and this information is passed to the lateral geniculate nucleus which decodes motion and depth perception (among other things), and then on to the visual cortex. The elemental stimulus of the visual cortex (what the neurons in V1 respond to) seems to be oriented line segments.

A good book for this is "Basic Vision", by Snowden, Thompson, and Troscianko.

I'm not sure if OP wants to know what I want to know.
a) Does the tree make 1-to-1 (or point to point) mapping on our retina and then on our brain? I assume, your answer will be 'yes', because there are lenses in our eyes and we observe image of a tree through lens in lab experiement.

b) Does a dog see a tree exactly what we see (exclude color)? If not, then our (all animals) vision is completely neuron-dependent?

 

[ebits21]

I'm not sure if OP wants to know what I want to know.
a) Does the tree make 1-to-1 (or point to point) mapping on our retina and then on our brain? I assume, your answer will be 'yes', because there are lenses in our eyes and we observe image of a tree through lens in lab experiement.

b) Does a dog see a tree exactly what we see (exclude color)? If not, then our (all animals) vision is completely neuron-dependent?

Well, the process is actually quite complicated. You will get a perfect, upside down and inverted, image on your retina given that you have proper focus (id est, you don't need glasses). However, the cells (rods and cones) that detect the image are not evenly spaced.

Most cone cells (that detect color) are concentrated at the fovea (the "central part" of your vision) and the rod cells (that detect black/white) are less dense and are concentrated in your peripheral vision. So the "grid" is not evenly spaced. In addition, everybody has a completely blind spot in their vision where the optic nerve leaves the retina (there's no receptors there). The reason you don't notice is because you brain basically ignores that there's a hole there at all.

Before leaving the retina, the cells of many receptors are passed on to bipolar cells and retinal ganglion cells. On average, a single retinal ganglion cell will connect to about 100 rod and cone cells; however, in the fovea they may connect to as little as five. This is why your vision, when looking straight ahead, is so much sharper than in the corner of your eye. The field of a ganglion cell is roughly circular, and if enough rods/cones fire in that field, the ganglion cell will increase its firing to the brain.

This is very simplified, but in the visual cortex of the brain particular areas process that incoming firing rate from the ganglion cells. Some of the first things to be processed are lines and edges in the image. From there, other areas are devoted to processing things such as color and movement. There's also a component that compares information from both eyes in order to create a 3d understanding of the World. After this low level processing other areas start to connect the information to concepts you have. So the green blob is finally interpreted as a leaf.

This visual field IS laid out as a sort of map in the visual cortex. It is, however, very distorted. Information from the fovea takes up much more space in the brain than other areas of the retina.

It's really impossible to say if what you see is the same as what other people see. Although, most likely it's fairly similar. A dog's vision would probably seem somewhat different. Humans and closely related primates have red receptors, whereas most other mammals, such as dogs, do not. This might be related to primates (including ourselves) being specialized to eat brightly colored fruit (The same reason why we need to make our own vitamin C, while dogs don't. In the jungle, fruit provided more than enough vitamin C so our bodies stopped making its own). Regardless, our perception of color is likely quite different than most other mammals. Our eyes also face forward with a lot of overlap between the visual fields, so we have better depth perception. Close-up, many things will appear blurry to a dog.

And of course vision is neuron dependent. You see with your brain, not your eyes.

For a bit more information you can have a look at these wikipedia articles:
http://en.wikipedia.org/wiki/Visual_system"
http://en.wikipedia.org/wiki/Retinal_ganglion_cell"
http://en.wikipedia.org/wiki/Blind_spot_(vision)" )
http://en.wikipedia.org/wiki/Opponent_process"
http://en.wikipedia.org/wiki/Trichromacy"

 

[Andy Resnick]

I'm not sure if OP wants to know what I want to know.
a) Does the tree make 1-to-1 (or point to point) mapping on our retina and then on our brain? I assume, your answer will be 'yes', because there are lenses in our eyes and we observe image of a tree through lens in lab experiement.

b) Does a dog see a tree exactly what we see (exclude color)? If not, then our (all animals) vision is completely neuron-dependent?

I can't give a definitive answer to (a), but AFAIK, the answer is 'no'- as in, you do not have a neuron dedicated to recognizing the tree in your front yard. That is different than asking if the eye creates an image of the tree on your retina (the answer is 'yes'- albeit an aberrated image). It's not clear at what point your brain recognizes and classifies the objects your retina senses.

I have no way of answering (b), other than simply comparing the optical characteristics of a dog eye versus a human eye. I'm reasonably sure the data is out there.

Don't forget, a single lens eyeball is only one solution to vision- compound eyes (insects) are another solution, and lens-free eyepits (snakes, in addition to 'regular' eyes) are another.

 

[ebits21]

I can't give a definitive answer to (a), but AFAIK, the answer is 'no'- as in, you do not have a neuron dedicated to recognizing the tree in your front yard. That is different than asking if the eye creates an image of the tree on your retina (the answer is 'yes'- albeit an aberrated image). It's not clear at what point your brain recognizes and classifies the objects your retina senses.

I have no way of answering (b), other than simply comparing the optical characteristics of a dog eye versus a human eye. I'm reasonably sure the data is out there.

Don't forget, a single lens eyeball is only one solution to vision- compound eyes (insects) are another solution, and lens-free eyepits (snakes, in addition to 'regular' eyes) are another.

The one cell = one concept idea is actually a fairly heavily debated issue in psychology/neuroscience literature and is known as the grandmother cell debate. The name comes from the idea that a single neuron in your brain represents your grandmother (http://en.wikipedia.org/wiki/Grandmother_cell" .

It does have its supporters (although I'm not one of them). Basically because particular neurons appear to only fire in response to particular concepts. I'm more of a supporter of the idea that patterns of neurons encode concepts and not single neurons.

As for when an image becomes a concept, this must certainly happen somewhere in the temporal or parietal cortex... the details of how this happens, however, is a mystery. A lot of the details of how things work in the brain are pretty poorly understood.