Science
Optical character recognition converts images into text
HERE’S THE INTERESTING PART
A scanned PDF is literally just a flat, dumb photograph of words. To make it editable, OCR algorithms aggressively scan the image, furiously looking for rigid shapes that remotely resemble letters. It essentially forces the computer to manually read the picture, letter by letter, frequently turning the word 'barn' into 'bum' by mistake.
One discovery leads to another.
Keep exploring facts in this topic.
Science →
