A short note to say thankyou to the DT team for DT’s OCR, which I’m finding is currently producing better results than (the more expensive) Acrobat Pro.
By way of example, here’s a text layer from the same PDF created from a TIFF file (not a great quality scan) with Acrobat Pro and DT:
(1) OCR’d using Acrobat Pro (downsampled to 300 dpi):
Treatmentis.!'otNa~nably necessaryendISnotrelatedtotheworklnJu,y.There1s
no evidence of any lnstabUltyon the MR scan
(2) OCR’d using Acrobat Pro (downsampled to 600 dpi, which AFAIK is the highest setting):
Treatment is .not masonably necessary and IS not related to the work fnJu,y. There 1s
no evidence of any lristabUltyo n the MR scan
(2) OCR’d using DT:
Treatment is not reasonably necessary and is not related to the work injury. There is no evidence of any instability on the MR scan
I’m not an expert and understand PDF files vary. I’m hoping that one day there will be an AI solution for cleaning up OCR errors which doesn’t involve sending information to third parties, but until then DT’s results are pretty good.