Rebuild guide
How to scan documents with your phone for a clean, usable PDF
Good recognition begins with a good source. A few deliberate capture choices can matter more than aggressive cleanup later, especially when the page contains small text, tables, stamps or faint 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. Messaging-app copies and screenshots often discard detail before recognition begins, so start from the original camera file whenever possible.
Use broad, even light
Bright indirect daylight or two lights from opposite sides usually produces a more even result than a phone flash. Move the page or lights until glare disappears from glossy paper. Watch for a hard shadow from the phone itself and for bright windows reflected in laminated documents.
Automatic exposure can make white paper grey or wash out faint print. Tap the page to focus, hold the phone steady for a moment and retake the image if small letters look soft when you zoom in.
Keep all four corners and hold the phone square
Position the camera parallel to the page rather than aiming from one side. Leave a narrow border around all four edges so cropping does not remove page numbers, handwritten notes or punched margins. Perspective correction can straighten a modest angle, but it cannot reliably restore detail that was never captured.
For multi-page work, keep the orientation, distance and lighting consistent. Consistency improves page sizing, text scale and later cross-page layout analysis.
- Include every corner and a small area beyond the paper.
- Check the top and bottom edges are close to parallel.
- Retake blur, glare, deep shadow or finger-covered corners immediately.
- Capture one complete page once; remove accidental duplicates before export.
Review each page before moving on
Zoom into the smallest important text, not just the page heading. Check names, reference numbers, decimal points, signatures, stamps and pencil marks. A full-page thumbnail can hide motion blur and clipped characters.
Keep the untouched capture even if you also create a cropped or enhanced version. Restoration choices such as thresholding, sharpening and background removal can erase faint evidence, so the processed scan should remain traceable to its source.
Build and check the PDF
Put pages in reading order, use a consistent page size and choose a compression level that leaves small characters legible. Open the finished PDF outside the scanning app, move through every page and confirm the orientation. If OCR is added, search for several difficult words and compare copied text with the image.
Do not delete the captures until the exported PDF opens correctly and has been backed up where you intend to keep it.
Camera access and privacy on the web
A browser camera tool requires HTTPS and explicit camera permission. Mobile browsers can also offer the phone's camera through a file-capture control. A professional scanner should explain permission failures, offer a manual fallback and store accepted pages safely before releasing the live camera frame.
Rebuild's public browser camera is still planned. The current web tools do not request camera permission, while the iOS product is in beta development and Android remains planned. Those availability labels will change only when a real public route or store release exists.