OCR text recognition (very) inaccurate in Acrobat Pro 2020 (text is PNG image "printed" to PDF)

Question

Is there a way to improve the OCR recognition of English text in Acrobat Pro 2020, when the text is a screenshot of a document (a PNG file), "printed" to a PDF, then run through OCR?

I am creating a study guide for a group reading this book, and I had hoped to save time transcribing the Table of Contents into the guide by screenshotting the e-book then running it through OCR. I screenshotted to a PNG, then "printed" to PDF (above); the OCR of this was very inaccurate - most of the words were not even English (many were not words in any language); many of the characters "read" were not even letters in the English alphabet. The language selected for OCR was English (US).

Any ideas will be welcome. Thanks!

gary_sc · Answer

Hi Matthew,

In a word, no. Standard minimum resolution for successful OCR is 300 ppi. Better quality OCR is about 600 ppi. Your screen is a whole lot less. At that resolution, you are getting about the average amount of success.

The content on the screen had to come from someplace. If you can get the original copy, you'll save yourself a lot of grief.

Sorry & good luck

Regístrese

Social Login

Bienvenido

Social Login

Escaneando el archivo en busca de virus

Este archivo no se puede descargar