Saw a pixeled url in a presentation. I attempted to read it by brute-forcing different options and seeing which one matches best.
With some brain power one can figure out the correct one given the results - but - it is bad at identifying "." so it won't give the right value.
Use create_t1.py to "train" (i.e match the font size, color, image size etc) against the given pixeled image (assuming the image starts with known text: "https://").
Then put the result in variable i of bf_t1.py and brute force the rest.
Note: This is a POC and horribly slow. It works if used correctly. Note: No AI here, only bruteforce with some measurments to identify if it's going the right direction.
- Don't pixel urls, just put a black box over them!
- Dont use python for things that need some speed