The internet celebrated net neturality the right way yesterday by clogging the tubes with heated discussion on some lady's dress color. What follows is an attempt to resolve this debate. Code can be found on github.
I took the original image and did some basic editing in GIMP to remove the background.
I processed each pixel to find its RGB value and then mapped the 100 most frequent RGBs to their nearest color (by name). Some of these RGBs ended up mapping to the same color. I combined these duplicates and sorted by overall frequency based on color name.
Here are the top 10 color names found and a sample corresponding hex value.
- Cool grey #8493be
- Light slate gray #737c9d
- Dark brown #5d4e2d
- Slate gray #717a9b
- Dark pastel blue #8c9bc6
- Dark lava #56462c
- Coffee #5f502f
- Glaucous #6d7ca7
- Raw umber #7b6c41
- Café noir #453827
The dress is dark brown and grayish-blue.
I am making the assumption that the person(s) who named these colors did a good job.