Scan Documents With Your Phone: A Searchable PDF Without an App

Good photos do half the work

No amount of processing rescues a bad photo. Lay the document flat on a dark, plain surface such as a wooden table, a dark folder or a piece of cardboard. The contrast between paper and background helps the automatic edge detection; on a white desk the sheet blends into its surroundings.

Take the photo in even daylight or under a bright ceiling light, but without the flash, which creates glare. Watch out for your own shadow and the phone's. Hold the phone as parallel to the sheet as you can and close enough for the paper to nearly fill the frame, while keeping all four corners visible.

For multi-page documents, take each page in the same position and light. Consistent photos produce consistent pages, and the PDF looks like one document rather than a collage.

• A dark, plain background
• Even light, no flash, no shadows
• Phone parallel to the sheet, all four corners in the frame

From photo to a straight page

In the tool, click Camera to photograph directly with your phone or webcam, or Upload to choose existing photos (JPG, PNG, WebP) or a PDF. Add page brings in more pages. Toolula looks for the paper edges in each photo and places four handles on the corners.

Check those handles. If they are not exactly on the corners, drag them into place, or use Re-detect or Use full image. The tool then corrects the perspective, turning a sheet photographed at an angle into a straight rectangle as if it had been on a scanner bed. Continue to filter & OCR takes you to the next step.

For documents with several pages, photograph them in reading order; each one appears as its own page with its own crop handles. A page that came out blurred or caught the wrong sheet can be removed and photographed again before you continue, so the finished PDF contains exactly the pages you need and nothing else. Front and back of an ID card or a two-sided form become consecutive pages of the same file.

Choosing the right filter

There are four filters. Enhanced Color whitens the paper and strengthens colours, which suits forms with coloured stamps. Original Color leaves the photo as it is, for example for photos or ID documents where every colour matters. Magic B&W Scanner produces crisp, high-contrast black-and-white text and suits plain text pages and receipts. Grayscale gives soft grey tones.

Shadow removal, contrast and rotation let you fine-tune the result. Pay particular attention to signatures and stamps: a strong black-and-white filter can swallow a light blue signature. When in doubt, use Enhanced Color for pages that carry a signature.

What OCR does, and what it does not

Without text recognition, a scan is just a picture: you cannot search for a word or copy a sentence out of it. OCR reads the text on the page and lays it as an invisible layer exactly over the image. The PDF looks like the scan but can be searched and selected. Toolula uses the Tesseract engine, compiled to WebAssembly and running in your browser tab.

Choose the document's language under OCR language. English, German, French, Spanish, Italian, Portuguese, Dutch, Turkish and Russian are available. The language pack is downloaded the first time and then cached in the browser; your document does not leave the device in the process. With the wrong language, accented characters and common words are recognised less reliably.

OCR is not flawless. Handwriting, very small print, blurry photos and tables with many lines lead to reading errors. For finding a name or an invoice number the result is usually good enough; if you process figures further, check them against the image.

Export, existing PDFs and privacy

Clicking Run OCR & export creates the file searchable-scan.pdf. The result shows how many pages, words and characters were recognised. If the count is suspiciously low, the photo was probably blurred or the language wrong; adjust the settings and run it again.

An existing image-only PDF can be made searchable too: upload it like a photo. Each page is brought in at 200 DPI and cropped to the full page, then filtered and recognised just like a camera scan. If a PDF already contains selectable text, it does not need OCR; the pages would otherwise become images with a new text layer and lose quality.

Edge detection, perspective correction, filtering and text recognition all run on your device. For ID documents, bank statements and medical paperwork, that is a real advantage over scanner apps that send images to their own servers.

Frequently asked questions