Make a scanned PDF searchable and check the recognized text
Updated · WaterPDF
Recognize printed English, German or French text in PDF scans on your device. Choose searchable PDF or TXT and check OCR errors before reusing the words.
No signup · Local file processing
Check whether you need recognition
Try selecting a sentence in a PDF reader. If the page is only a picture, ordinary text extraction has no words to copy. OCR PDF recognizes printed words from the page image. If the words are already selectable and correct, PDF to text can extract them without another recognition pass. Keep the original scan so you can compare the result.
Prepare an upright, readable scan
Use a clear scan with the whole line visible. Rotate sideways pages first; faint print, skew, shadows, handwriting and very small text can reduce accuracy. For a JPG or PNG, create a PDF with Images to PDF first. OCR supports up to 10 PDF pages and 100 MB input, with up to 8 megapixels per page and 80 MB of combined rendered images. Long processing stops after five minutes; split longer documents before recognition.
Choose the language used on the page
Choose English, German or French to match the printed words. The selected language helps recognize its vocabulary and characters, but it does not guarantee spelling or punctuation. A page with several languages or unusual technical terms can need more correction. The language assets and recognition engine run on your device, with no document upload.
Choose searchable PDF or plain text
Searchable PDF keeps each rendered page as a picture and adds a recognized text layer for searching and copying. It does not turn the scan into original editable paragraphs or cells. TXT gives you the recognized plain text. Interactive forms, links, bookmarks and certificate signatures from the original are not retained as interactive objects in the rebuilt searchable PDF.
Review names, amounts and reading order
Search for a few known words and copy a short passage into a plain text editor. Compare names, dates, decimal separators, reference numbers and similar characters such as 0 and O or 1 and I against the image. Review multi-column reading order and every important table value. A searchable download can still contain recognition mistakes; correct reused text before relying on it.
Continue to Word, Excel or text extraction
After checking the searchable PDF, try PDF to Word, PDF to Excel or PDF to text. These converters use the resulting selectable text and do not run OCR automatically. OCR does not recover original spreadsheet formulas or reliable table boundaries: a scanned table is still an image, so inferred rows and columns can need correction. Use the scan alongside the converted file.
Choose the right route for the page
| What you have | First tool | What to review |
|---|---|---|
| PDF with correct selectable words | PDF to text | Reading order and missing text |
| Printed scan with no selectable words | OCR PDF | Recognition errors and page order |
| Photo in JPG or PNG | Images to PDF, then OCR PDF | Orientation, clarity and recognized text |
| Scanned table to reuse in Excel | OCR PDF, then PDF to Excel | Every amount, blank cell and inferred column |
OCR is a recognition aid. Preserve the scan and check important information against it.