Hi, i’m working one the captcha level 1 and i try to make a kind of OCR from scratch with canvas in javascript.
I will transform a pixel matrice in binary number (0 -> black, 1->green) and compare it with a kind of custom map_list like binary_number -> character.
I know it will probably work but the question is for the nexts levels, is the pixels matrix way a good idea ? When a picture is rotated, it’s way more difficult to work directly on pixels, so i should more work on the general shape of character than the matrix, right ? How can i do that ? I have no idea how to start.
Tks for help, sorry for this bad english.
