Ooh, ooh, I know! You should ask a computer.
Take a photograph with the highest-fidelity camera you can find under whatever lighting conditions you consider favorable and move the image to Photoshop (or similar). Clip a section of iris and put it in another canvas so that it entirely fills the frame. Scale, not crop, the image down to one pixel, and behold! The lowest-effort algorithmic measurement of your eye color has been revealed.
The idea is to use a program's
rasterization function to convert a many-pixel image into a single-pixel image. This is analogous to what someone might say your eye color is if they saw it from some distance, and thus it is a reasonable measure of one's apparent eye color.
In order to prove the image scaling method is actually averaging pixels instead of sampling just a subset of them, you can try making a 4-pixel image of two red and two blue pixels and scaling it to one; you should get a pretty purple pixel with RGB color values precisely the average of the red and blue pixels' values.
Pixels in each frame except the two with iris details in them are all scaled up 1600%.
Photoshop says the red is R:255 G:0 B:0 and the blue is R:54 G:0 B:255.
The purple reads as R:155 G:0 B:128.
(255 + 54) / 2 = 154.5, and (255 + 0) / 2 = 127.5 ... so the approach worked!
Iris is from WikimediaThere are some drawbacks to this approach, but they're all consequences of assuming the complex surface of an iris can be summed up as one single color. Asking the computer what your eye color is directly answers that specific question for a given photograph. To improve the accuracy, try using four squares of iris: one each from above, below, left and right of the pupil, aligned into a 2x2 grid in the new image.
Author
Topic: How do I figure out my eye color? (Read 41776 times)