If we do the
algorithm of distribution of attention in the image , then we need to do according to the mind, which means that it becomes vital to determine the “power of color”.
Color perception
The sensation of color
depends on the signal entering the brain from S, M and L cones. At the same time, some colors determined by a person feel several types of cones at once, and it is because of this that some colors appear brighter than others that are similar in tone and saturation.
one-response')
It is interesting, but
what would the colors look like if the power of their perception were the same with the saturation? Answering this question, I wrote a program illustrating the relationship between color and perception. And here are the results of her work:
Color target: original , post-processing .
Other color target: original , post-processing .But the above images show the result only for full color vision. I wonder
how people look at the world of color blindness?Color blindness
And color blindness can be different: color blindness arises mainly due to the lack of certain cones. In case of shortage of two types of cones, the colors do not differ at all, the world looks monochromatic, as in a black and white photo. But in the event of a shortage of cones of one of the types S, M or L, a situation of indistinguishability of certain colors arises.
Let's try to simulate, based on the data of the color sensitivity of cones, the perception of color blind!
Frankly, these models are not calibrated. Ideally, a color blind should not feel a significant difference between the original and the result of the simulation. But in order to imagine how curious? :)
The pictures - images after the normalization of the power of color perception. Targets and screenshots are mine, and photos are taken from sites
from here ,
from here and
from here .
Originals:




Lack of L cones (indistinguishability red):





Deficiency of M cones (indiscernibility of green):





Lack S cones (not visible blue and purple):





In my opinion, people with indistinguishability of blue shades in everyday life have the easiest way. Much harder for nondiscriminating red. Traffic lights, road signs, warnings are often invisible.
PS
Having received several requests to show the program after writing the post about modeling the view , I decided to make a simple web service like “fill the picture> get the result”. In this regard, began to rewrite the program that generates the dynamics. I accept suggestions how it should look.