I am implementing an OCR system for a final year project. An initial step is to be able to 'chop' up an image of text by a bounding box, then line by line, word by word, until I have a set of individual glyphs for my system to recognise each character.

Can AWT be used to access the pixels of an image in this way?