This directory (segmenter/*) has the code to cut up an aligned page image into individual characters.

A proper readme for the project is in the directory above.

Our segmentation is specific to lineprinter listings of fixed pitch (aka monospaced) text.

If you need to OCR variable-width text with our code, maybe use https://github.com/SAmohsen/CharacterSegmentation_OCR
to do the segentation instead.