PdfRecast guide
Recognize text in scanned PDF pages
Optical character recognition turns visible letter shapes into machine-readable text. PdfRecast runs OCR for supported pages in the browser and exposes the detected regions so users can review and correct them instead of silently accepting uncertain output.
How to use this workflow
- Open a scanned PDF and confirm the page is upright and readable.
- Start OCR for the required pages and monitor recognition progress.
- Inspect the detected text boxes against the scan, correcting errors and missed text.
- Save the reviewed document and test search or selection in the output.
Recognition limits
Accuracy varies with language, character set, scan resolution, compression artifacts, skew, handwriting, columns, tables, and background noise. A confidence score is not proof that a word is correct.
OCR does not establish that the source document is authentic, nor does it validate the meaning of the extracted text. Preserve the scan and compare critical content manually.