@0y3ahSansAcut3 OCR fonts are an interesting exercise in trying to distinguish similar things. Not just 0/O and 1/I... a smeary print or scanning too fast can make a 5 look like an S or vice versa, a G like a Q or a 6.
Altering the profile of the letter is one way to help with this. Train the device to recognize an O is diamond-shaped and a 0 is square. (Which btw is kind of entertaining if you're a typographer and used to seeing the Oh being rounder and zero being skinnier).
Tangentially this is relevant to periodic attempts at designing fonts for dyslexic readers. One of the techniques that help people with dyslexia is for similar-ish characters (C, G, 0, O, Q, or I, 1, l, |, /, !) to be drawn more distinctly, and to have symmetrical characters (I, T, A, U, etc.) drawn asymmetrically. It turns out that one of the better fonts for some dyslexic readers is Comic Sans.
Altering the profile of the letter is one way to help with this. Train the device to recognize an O is diamond-shaped and a 0 is square. (Which btw is kind of entertaining if you're a typographer and used to seeing the Oh being rounder and zero being skinnier).
Tangentially this is relevant to periodic attempts at designing fonts for dyslexic readers. One of the techniques that help people with dyslexia is for similar-ish characters (C, G, 0, O, Q, or I, 1, l, |, /, !) to be drawn more distinctly, and to have symmetrical characters (I, T, A, U, etc.) drawn asymmetrically. It turns out that one of the better fonts for some dyslexic readers is Comic Sans.