Stare at the colored dots on the girl’s nose in the photo above for 30 seconds. Then look at a white surface (blank browser, perhaps) and start blinking. You should see a non-negative image of the girl.
One study provides the explanation for how these negative “afterimages” are produced.
Negative afterimages are caused when the eye’s photoreceptors, primarily those known as cone cells, adapt from the overstimulation and lose sensitivity. Normally the eye deals with this problem by rapidly moving small amounts , the motion later being “filtered out” so it is not noticeable. You can test this yourself when you lay in bed at night. If you turn out the lights and stare into a dark corner of the room, everything will begin to go black. This is your visual system getting “tired” and losing sensitivity. Only if you look at a different area of the room will things return to normal.
However, if a color image is large enough that the small movements trying to keep sensitivity are not enough to change the color under one area of the retina, those cones will eventually tire or adapt and stop responding. The rod cells can also be affected by this.
When the eyes are then diverted to a blank space, like the white space after staring at the image above, the adapted photoreceptors send out a weak signal and those colors remain muted. However, the surrounding cones that were not being excited by that color are still “fresh”, and send out a strong signal. The signal is exactly the same as if looking at the opposite color, which is how the brain interprets it.
Ewald Hering explained how the brain sees afterimages, in terms of three pairs of primary colors. This opponent process theory states that the human visual system interprets color information by processing signals from cones and rods in an antagonistic manner. The opponent color theory suggests that there are three opponent channels: red versus cyan, blue versus yellow, and black versus white. Responses to one color of an opponent channel are antagonistic to those to the other color. Therefore, a green image will produce a magenta afterimage. The green color tires out the green photoreceptors, so they produce a weaker signal. Anything resulting in less green, is interpreted as its paired primary color, which is magenta.
This is the reason that looking at the three colored dots on the girl’s nose produce the afterimage which is colored “opposite” (following the theory) to those dots.