Question

I want to identify the colors of this color checker chart in a little robust way. In the first step not every color has to be detected correctly.

I ectracted already the subimages, so that I have squares with only one color.

The idea for detecting the color (and print them out) was to transform the colors from RGB to the HSV color model.

Does anybody has a better solution or can help? Best regards!

Was it helpful?

Solution

I think you want an algorithm like the following:

  • Compute the mean RGB value for a given subimg.
  • Compute the Euclidean distance from your mean RGB value to the RGB value in each of the squares.
  • Return the square name that is closest to your RGB value.
Licensed under: CC-BY-SA with attribution
Not affiliated with StackOverflow
scroll top