When using OCR on scanned documents, hand drawn images are confused for text and are reliably ruining the output PDFs. The easy fix would be to allow users to exclude "zones" (boxes) of the document from OCR scanning, OR allow users better tools to remove zones of detected text.

As of current, there isn't any easy way of correcting OCR from scanning hundreds of nonexistent characters from drawings and images. If OCR has detected many characters on seperate lines, the user will need to painstakingly shift through each correction until the document is cleared of all errors.

I'm quite confused why this isn't already a feature on a paid application such as this one, especially one that comes from the company who makes Photoshop.