Mono-Match-Mystic

Matthew Merrill 2021

Can't find the icon shared by two pentagons? Let your computer find it for you!

Requires pillow and numpy to be installed. Pillow is used for image manipulation and opening the cards in a pop-up window to display the results. Numpy is used for calculating the Mean-Squared Error of two images.

Find match across two cards

python3 main.py find_match path/to/card1.png path/to/card2.png

Find all matches shared across a set of cards

python3 main.py find_matches path/to/card1.png path/to/card2.png path/to/card3.png

Why?

To solve the card game made by Matt Parker as part of his "The Maths Of Dobble" video! Patreons were sent cards to solve, and my set of cards proved difficult to verify. I figured writing a program would be faster (and maybe a bit more fun).

How does it work?

Each icon is extracted from the original image then blurred by rotating it in a circle and blending all the rotations.

Then, the blurred icons are compared against each other in N^2, brute force glory. The two blurred icons with the lowest Mean-Squared Error are very likely the matching icons. For my cards, this worked perfectly!

How does the computer know where the icons are inside the image?

A combination of the locations being shown in Matt Parker's "making of" video, a keyboard, and determination.