As part of a challenge in the Damncon 2022 CTF, I had to decode the follownig image:
This is clearly the Hexahue Cipher. However, not without surprise, I could not find any automatic image decoder online. All the decoders (like dcode.fr) let you type your message using the images, which is perfect for small messages but not applicable to this image. I hence decided to go ahed and write a HexaHue decoder by myself.
Cropping the image
First of all, I had to crop the original image into small pieces, each one representing one hexahue char. This is an easy task for image manipulation libraries like PIL. I only had to compute the average size of a block and the average distance between blocks. Then I removed 3 extra pixel from each side to be sure not to include the background. Finally, I saved the blocks in numbered images in the same folder.
Image to Code
Next step was to parse the images. To do so, I took the average color value on 4x4 pixel blocks from top, middle and bottom (both left and right) of the image. The average on each RGB component was then passed to the function roundcolor and rounded to the closest value among 0, 127 and 255. This is an easy way to prevent mistakes, since those three are the only possible RGB values for hexahue encoding. Each color was finally mapped to a color code, like R for red, G for green and so on. Due to the high number of G in color names, I used D for black and X for gray. So each box is translated into a 6 digit string based on its colors, starting from the top left.
Code to string
Finally, I wrote a dictionary to map each color combination to the corresponding letter.
The full solution script can be found at hexahue.py