A phone photograph may have shadows or tilted edges. Start with a readable capture, then adjust corners and colour mode. A white-looking page is not proof that all its text survived.
Retake or process?
Our recommendation is to retake photos with glare covering text, missed focus or cropped paper edges. If the source is unreadable, do not expect a filter to reliably reconstruct it.
Tesseract documents recognition problems from unevenly dark backgrounds and skew. Preparing the image matters when OCR follows, but processing does not guarantee correct recognised text.
Set the page corners
Open Scan to PDF and select document photos. Move the four corner points to match the paper. The tool transforms the image using your chosen corners; avoid excluding names, dates or page numbers.
Check each page independently. Rotate an upside-down photo and recheck its corners. One correctly framed page does not establish that the whole batch is correct.
Choose a mode by the content
| Mode | Our suggested starting use |
|---|---|
| Colour | Stamps, highlights or illustrations where colour matters |
| Grayscale | Reduce colour distraction while retaining intensity differences |
| Black and white | Dark text on paper, with extra checks for thin or faint marks |
The site's black-and-white filter compares local brightness to separate marks from the background. It may not suit faint signatures or pale colours. Compare with colour before choosing it.
Inspect the saved PDF
- Check page count and order.
- Zoom into names, numbers, punctuation and footer details.
- Look for missing stamps or faint text.
- If black and white removed details, regenerate using colour or grayscale.
This step produces a PDF of page images; it does not automatically add searchable text. Run OCR PDF separately if you need searching or copying, then verify the recognised words. See Thai PDF OCR.
Keep the original photographs. If the checked PDF is too large, use Compress PDF on a copy and reopen it after compression.
