How to use it
- Choose a scanned PDF of 20 pages or fewer.
- Pick the language closest to the page.
- Download the Word file and the text file, then proofread names and numbers.
A scan is a picture of a page, so normal PDF-to-Word finds nothing to copy. OCR looks at that picture and writes the characters it can see. English, Hindi, Spanish, French, and German recognition are offered.
The first run can download language data from a public network address. That download is the recognition model, not your PDF. The document stays in the tab. OCR makes mistakes on handwriting, stamps, and tiny print, so read the result before you rely on it.
Questions
Why is there a network request?
The recognition engine and the language file are fetched once. SKY13.app does not receive the PDF.
Can I OCR 100 pages?
This page stops at 20 so a phone can finish. Split the scan and run each part.