How does human eye decode different geometric shapes such as circle?

[PainterGuy]

Hi,

How does a human eye classify any shape as a circle, square, triangle etc.?

Let's focus on a circular shape. Suppose we have a circle drawn in white on a black surface. The light falls on the retinal cells. I think the light falling on the retina will constitute a circular shape as well, and hence will activate the cells along the circular boundary. Is it the activation of the cells which signals or helps the brain to decode it as a circular shape?

Thanks for the help, in advance!

 

[BillTre]

Try looking at this.

Receptor cells in the retina detect light and signal this to neurons in the retina.
As the signals work their way along the visual paths to and in the brain, signals are passed along, but can processed to detect activity between neighboring retinal cells.
There are many of these working in parallel for each little patch of retina, for each little patch of the visual field.
This allows for visual feature detection to be built up through the detection of simple patterns like a circle surround of activation in the middle combined with a surround of inhibition (on a much smaller scale than your projected circle).
A row of activated centers in these center surrounds can be detected by activating line detector neurons.
A circle would be like a line but curving.

Some processing occurs in the retina, but more goes on in a number of visual centers in the brain.
The details get very complicated.

 

[symbolipoint]

BillTre, I do not follow too well what you said, but I take you are saying, your eyes can do some of the thinking?

 

[BillTre]

BillTre, I do not follow too well what you said, but I take you are saying, your eyes can do some of the thinking?

Some visual processing occurs in the eyes. There are about three layers of neurons directly underneath (on top of them in vertebrates actually) the photoreceptors in the retina.

However, the retina is actually embryologically derived from parts that would otherwise form brain tissue. This is why some people say the retina (and photoreceptors) is part of the brain. For this discussion, this is not relevant to eyes vs, brain, but I include this in case someone hears about it.

How much processing occurs in the eyes vs. the brain differs in different animals. In people, I believe it is mostly in the brain.

 

[PainterGuy]

Thanks a lot for the help.

It wouldn't be wrong to say that it still starts with the pattern(s) created by the incoming light on the receptor cells in retina. The decoding/processing and pattern recognition parts could be done in different places such as brain.

I think the system could be fooled as well. Let me elaborate.

Suppose one is in a dark room and one of the walls show the circular pattern shown below.

The white circle creates the pattern as shown below. The green is pattern created on the retina receptor cells and some of those receptors have been labelled for clarity. The orange circle represents the retina surface with all its receptor cells.

Now if the all the retina receptors cells have been compressed along horizontal axis as shown below. The pattern on the wall is an ellipse instead of a circle. If the same set of receptors are activated as shown below, the brain would still recognize the pattern as a circle.

I hope it makes little sense.

 

[pinball1970]

Thanks a lot for the help.

It wouldn't be wrong to say that it still starts with the pattern(s) created by the incoming light on the receptor cells in retina. The decoding/processing and pattern recognition parts could be done in different places such as brain.

I think the system could be fooled as well. Let me elaborate.

Suppose one is in a dark room and one of the walls show the circular pattern shown below.

View attachment 303366

The white circle creates the pattern as shown below. The green is pattern created on the retina receptor cells and some of those receptors have been labelled for clarity. The orange circle represents the retina surface with all its receptor cells.


View attachment 303367

Now if the all the retina receptors cells have been compressed along horizontal axis as shown below. The pattern on the wall is an ellipse instead of a circle. If the same set of receptors are activated as shown below, the brain would still recognize the pattern as a circle.

View attachment 303368

I hope it makes little sense.

How would those cells be compressed like that? That would equate to some sort of significant trauma to head and seeing an ellipse in stead of a circle would probably be low on your priorities.

EDIT: I had forgotten about the below.
I was imagining transitioning from normal shape to more ellipsoid which would involve some sort of trauma.
A head in a vice type scenario

 

[pinball1970]

Thanks a lot for the help.

It wouldn't be wrong to say that it still starts with the pattern(s) created by the incoming light on the receptor cells in retina. The decoding/processing and pattern recognition parts could be done in different places such as brain.

I think the system could be fooled as well. Let me elaborate.

Suppose one is in a dark room and one of the walls show the circular pattern shown below.

View attachment 303366

The white circle creates the pattern as shown below. The green is pattern created on the retina receptor cells and some of those receptors have been labelled for clarity. The orange circle represents the retina surface with all its receptor cells.


View attachment 303367

Now if the all the retina receptors cells have been compressed along horizontal axis as shown below. The pattern on the wall is an ellipse instead of a circle. If the same set of receptors are activated as shown below, the brain would still recognize the pattern as a circle.

View attachment 303368

I hope it makes little sense.

Actually you can get what is called a stigmatism where the eye is more egg shaped, elongated.
This can lead to vision issues which can be corrected

 

[BillTre]

I think the system could be fooled as well.

This would be optical illusions. There are lots of them. They affect different steps in the process of building up visual perceptions.
https://newyork.museumofillusions.us/three-types-of-optical-illusions-museum-of-illusions/

They are based on a misrepresentation of the visual world in our visual perceptions.

Now if the all the retina receptors cells have been compressed along horizontal axis as shown below. The pattern on the wall is an ellipse instead of a circle. If the same set of receptors are activated as shown below, the brain would still recognize the pattern as a circle.

Not exactly sure of your point, but the brain can make up for some distortions.

There are more (a higher density per area) photo-receptors at the retinal fovea (basically the center). There are also more color photo-receptors there (vs. non-color discriminating).
The brain constructs the visual scene you are aware of and "hiides" some expected problems that might arise from these non-linearities.