What is the resolution of the human eye in megapixels? It is difficult to compare resolution of the human eye with digital camera resolution because the concepts are different.
If you had to pick a number, you could say that the human eye has 1 million “pixels” of different sizes (tiny high-res pixels at the center, large low-res pixels in the periphery). At the center of the visual field, the resolution might be equivalent to a 10 megapixel
camera. There are 100 million individual photoreceptors in the retina!
The "resolution" of the human eye and the Hubble telescope (the ability to resolve detail) is measured in angular resolution or angular degrees per "pixel". Angular resolution varies from 1 arc minute (1/60 of an angular degree) at the fovea to possibly 1 degree or more in the periphery. Note that the fovea is the center-most region of vision. It is around 2x the size of your thumb with arm outstretched or a bit bigger than the moon in the sky.
Here is a depiction of solid angle (in angular degrees) with the eye at the center: For a digital camera, the angular resolution varies quite a bit according to the zoom lens factor coupled with the pixel density of the CCD.
In a digital camera, the "resolution" refers to the total number of pixels on the imaging plate (the CCD) measured in megapixels. So the analogy for humans might be the total number of number of photoreceptors ("pixels"?) in the retina, which is around 100 million. Or it could be the total number of nerve fibers sent from the eye to the brain, which is about 1 million. Most of these "pixels" come from the fovea.
One could argue based on angular resolution and the typical viewing distance of photographs, that that the "megapixel equivalent" of the eye is around 10 megapixels at the fovea and 0.1 megapixel in the periphery.
To continue the resolution confusion, in print media, resolution refers to pixel density in dots per inch (or other unit) on a surface, e.g. 72 or 300 dpi. The resolution of an inkjet printer is the highest pixel density it can print. For computer monitors, there is both resolution (horizontal and vertical pixel count) and pixel density or dot pitch, measured in either pixels-per-inch, e.g. 120 PPI, or millimeters per pixel, e.g. .20mm. Dot pitch would be the most analogous to the human eye measure of angular resolution if one assumes a fixed reading distance. The pixel density of the human eye at a 20" reading distance might be around 170 dpi or a .14mm dot pitch at the fovea and 2 dpi or 15 mm in the periphery.
Thanks for A2A. Dave already gave a fabulous answer, read his first or instead. However I try whenever possible to answer A2A so here goes.
I am a fundamental theoretical math/physics bounds kinda guy.
No one seems to have brought up this angle in an answer yet (pun intended!) so here goes!
Behind the Cornea is the eye's lens. Together they focuses light using Snell's law. The effective size of the human lens is ~1cm=10mm, as has been pointed out in other Quora answers. The sharpest (narrowest wavelength) light we see is blue(ish), with wavelength ~500nm. Diffraction theory, which is a consequence of the quantum properties of a photon, leads to the Airy disk as the bound of possible resolution. This is often called the Rayleigh diffraction limit, see Angular resolution. It is (roughly) the lens width divided by wavelength. For us that is 20,000 in each dimension or 400 Million. Digital cameras sometimes have more pixels than the resolution limit of the lens to reduce pixelated distortion, see Image resolution. This explains the different answer (slightly) that Dave provides.
I am a fundamental theoretical math/physics bounds kinda guy.
No one seems to have brought up this angle in an answer yet (pun intended!) so here goes!
Behind the Cornea is the eye's lens. Together they focuses light using Snell's law. The effective size of the human lens is ~1cm=10mm, as has been pointed out in other Quora answers. The sharpest (narrowest wavelength) light we see is blue(ish), with wavelength ~500nm. Diffraction theory, which is a consequence of the quantum properties of a photon, leads to the Airy disk as the bound of possible resolution. This is often called the Rayleigh diffraction limit, see Angular resolution. It is (roughly) the lens width divided by wavelength. For us that is 20,000 in each dimension or 400 Million. Digital cameras sometimes have more pixels than the resolution limit of the lens to reduce pixelated distortion, see Image resolution. This explains the different answer (slightly) that Dave provides.
I was interested in how the human eye compares to digital imaging and so far have found and compiled this table from the sources listed at bottom:
Human Eye Specifications (typical):
Human Eye Specifications (typical):
- Sensor (Retina) : 22mm diameter x 0.5mm thick (section); 10 layers
- Resolution : 576MP equiv.
- Visual Acuity : ~ 74 MP (Megapixels) (printed) to show detail at the limits of human visual acuity
- ISO : 1 - 800 equivalent
- Data Rate : 500,000 bits per second without colour or around 600,000 bits per second including colour.
- Lens : 2 lenses - 16mm & 24mm diameter
- Dynamic Range - Static : contrast ratio of around 100:1 (about 6 1/2 f-stops) (4 seconds)
- Dynamic Range - Dynamic : contrast ratio of about 1,000,000:1 (about 20 f-stops) (30 minutes)
- Focal Length : ~ 3.2mm - (~ 22mm 35mm equiv)
- Aperture : f2.1 - f8.3 (f3.5 dark-adapted is claimed by the astronomical community)
- FOV Field of View : 95° Out, 75° Down, 60° In, 60° Up
- Color Space - 3D (non-linear) RGB
- Color Sensitivity : 10,000,000 (ten million)
- Color Range : 380 to 740 nm
- White Balance : Automatic (constant perceived color under different lighting)
- Refresh Rate : foveal vision (high-quality telescopic) - 3-4fps; peripheral vision (very inaccurate) - up to 90fps
Obviously compiling this creates some very big assumptions - many of these are interrelated to the brain's processing of the eye's signals. I am sure that while the above statistics are factual, presenting them as a direct apples-for-apples correlation to common camera specifications is probably fraught with inconsistencies. So bear in mind that I did this just for fun!
Enjoy
Enjoy

0 comments:
Post a Comment
Note: only a member of this blog may post a comment.