Scan documents with your phone
Capture quality limits what OCR and compression can recover. A flat page, even light and a square camera angle retain small text, rules, stamps and handwriting.
Prepare the page and camera
Place the document on a flat, contrasting surface and smooth folds without damaging the page. Clean the camera lens, use the main rear camera rather than a lower-quality front camera, and avoid digital zoom. If the document is bound, support the facing pages so the text near the spine is not curved or shadowed.
Capture at the camera's normal resolution and keep the original camera file. Messaging-app copies and screenshots can discard detail before recognition begins.
Use even lighting
Use bright indirect daylight or two lights from opposite sides. Move the page or lights until glare disappears, and check for shadows from the phone or reflections in glossy and laminated pages.
Tap the page to focus and hold the phone steady. Retake the image if small type looks soft, white paper turns grey or faint print disappears.
Frame the whole page
Hold the camera parallel to the page and leave a narrow border around all four edges. Perspective correction can straighten an angled page, but it cannot restore clipped page numbers, notes or margins.
For multiple pages, keep the orientation, distance and lighting consistent. This keeps page size and text scale consistent in the PDF.
- Include every corner and a small area beyond the paper.
- Check that the top and bottom edges are close to parallel.
- Retake pages with blur, glare, deep shadow or covered corners.
- Capture each page once and remove accidental duplicates before export.
Inspect each capture
Zoom into the smallest important text. Check names, reference numbers, decimal points, signatures, stamps and pencil marks; a page thumbnail can hide blur and clipped characters.
Keep the original capture alongside cropped or enhanced versions. Thresholding, sharpening and background removal can erase faint marks.
Test one representative page first
Before scanning a long packet, choose a page with the smallest type, table rules, a stamp or handwriting. Build and reopen a one-page PDF, zoom into those details and test OCR if searchable text is required. Adjust the light, distance or camera position before repeating the setup for the full batch.
Write down the expected page count and the order before you start. A quick first-page test is cheaper than discovering after twenty captures that a shadow hid every footer or the receiving system rejects the chosen file size.
Assemble the PDF
Put pages in reading order, use a consistent page size and choose compression that leaves small characters legible. Open the PDF outside the scanning app and check every page and orientation. If OCR was added, search for difficult words and compare copied text with the page image.
Keep the captures until the PDF opens correctly and is stored where you intend to keep it.
Official sources
Sources checked on 14 August 2026 for the technical and procedural details cited in this guide.