# How many mega pixels does the eye have?

What is the number of mega pixels available in the human eye? It seems that newer camera models continuously keep increasing their pixel count. However, they never seem to be capable of reproducing what the eyes can see.

• According to calculations done on this it is around 576 MP Nov 14 '15 at 11:13
• There is a similar question on Photography.SE photo.stackexchange.com/q/39389
– user19750
Nov 14 '15 at 11:21
• @Ghost A quote from my late answer on that question: "None of these calculations are scientifically accepted." This still stands and the question is purely subjective and speculative. Nov 16 '15 at 3:33
• @GoodGravy - while the subjective conscious perceptual representation of the visual field cannot be expressed in megapixels, the retinal composition of photoreceptors and their relative contributions to the image can be approximated. It is not purely subjective and speculative at the retinal level.
– AliceD
Nov 19 '15 at 12:55

The total number of photoreceptors ('pixels') in the human retina is 63 million. Approximately 3 million of them transmit focused, color-coded information.

Background
In essence, each photoreceptor on the retina acts as a photosensitive sensor. We have rods and cones in our retinae. Rods confer gray scale vision, while cones are responsible for color vision.

The raw average tally of rods in the human retina is 60 million, and cones count a little over 3 million (Jonas et al, 1992). Hence, we could conclude we have 63 megapixels per eye available.

However, this does not mean that 63 million pixels are sent off to the brain. Instead, just as with a JPG image, it's instantly reduced in size even before the information leaves the retina.

Most importantly, rods are specialized for scotopic (night) vision and 10 to 30 rods connect to a single bipolar cell (Kolb, 2011). Hence, this feat only reduces the number of 'rod pixels' to about 3 million. Cones generally synapse onto a single bipolar. Hence, 6 megapixels leave the retina approximately.

The retinal image is generally focused onto the fovea, which is the central 18o of the total ~160o that is covered by a retina (Fig. 1). The remainder of the retina may receive a blurred image projection.

Fig. 1. Image focussing through accommodation on the retina. Source: Health Problem and Solutions

The fovea is densely packed with cones and contains nearly all of them (fig. 2).

Fig. 2. Distribution of rods and cones in human. Source: (Kolb, 2011)

Hence, the focused part of the image may be encoded by as little as 3 million cones. Since rods don't provide color information, it can be concluded that focused, color-coded information is conveyed by 3 million photoreceptors per retina.

Ultimately, however, it is important to note that the brain fills in missing gaps. For example, the optic disc, i.e. the point where the optic nerve leaves the eye, measures about 1.5 mm in diameter (Kolb, 2011). Given that the total diameter of the retina is about 30-40 mm this is quite a big blind spot. The brain simply fills in that gap by estimating what supposed to be there. In fact, many people with degenerative diseases of the retina don't even notice their deficits until they start bumping into objects that they did not notice. However, they do not see black holes, but instead their brain has been filling in the gaps with the information available in the remaining image, just like it does with the optic disc in normally-seeing people.

References
- Jonas et al, Graefes Arch Clin Exp Ophthalmol (1992); 230(6):505-10
- Kolb, Webvision, The Organization of the Retina and Visual System, Simple Anatomy of the Retina (2011)
- Kolb, Webvision, The Organization of the Retina and Visual System, Circuitry for Rod Signals Through The Retina (2011)

Actually the human eye has variable resolution, with higher resolution around the fovea and lower as we move away from it. So I don't think the 576 MP is correct.

Our eyes are really horrible in many optical properties. The brain makes up for these errors and corrects everything, so in this sense your question is a poorly proposed one, as we should probably perhaps stop improving resolution of camera sensors and start improving the software that analyses the data they receive.

• +1 for a reasonable answer, but can you back this up with citations? Nov 15 '15 at 15:06