When you scan a document into a PDF, the text on the page is actually a photo. That's why you can't select, search or copy it. OCR (Optical Character Recognition) turns that image into real text.
What exactly does OCR do?
OCR recognizes the letters and words in the page image and converts them into editable, searchable text. So you can summarize, search within, or copy a scanned contract, deed or invoice.
OCR step by step
- Upload the scanned PDF: Add your photo/scan-based PDF to the OCR tool.
- Convert to text: OCR turns the text in the page images into real text on your device (Turkish + English).
- Use the text: Now search, copy, or feed it into summarize or translate tools.
Scanned PDF → Text (OCR) — Turn text in a scanned document into real text.
Processing happens entirely on your device — your document is not uploaded. Best results come from clear, high-resolution scans.
What determines OCR quality?
OCR accuracy depends heavily on the input. The same engine is near-perfect on a clean scan and confuses letters on a shadowed phone photo. The main factors are:
- Resolution: scans around 300 DPI perform noticeably better; below 150 DPI the error rate climbs fast.
- Contrast: faded ink or a grey photocopy makes letters hard to separate from the background.
- Skew: even a few degrees of tilt disrupts line tracking.
- Typeface: standard print faces are recognized well; handwriting and decorative fonts are not reliable.
- Layout: reading order can get scrambled on multi-column pages and tables.
Accented and non-English characters
Characters outside the basic Latin set are where OCR slips most often — Turkish dotless ı against i, or capital İ against I, is a classic case. On documents where a single character matters, such as identity papers, deeds or contracts, verify the output by eye. If you only need the text for searching later, small errors are harmless; if you will copy it into something official, checking is essential.
Searchable PDF or plain text?
These are two different needs that often get confused. To keep the document's appearance and search inside it, an invisible text layer is placed over the page: the document looks identical but Ctrl+F works. To move the text elsewhere, a plain-text output serves better, but the page layout is lost. For most people the searchable PDF is more useful because the document keeps its official look.
Preparing the document before OCR
A little preparation raises recognition noticeably. Straighten skewed pages with the rotate tool, delete blank or duplicated pages, and run phone photos through the Document Scanner first. Because the scanner finds the edges and corrects perspective, OCR receives a far cleaner page.
What you can do after OCR
- Export the document to Word and edit it.
- Have AI summarize a long contract.
- Pull amounts and dates from an invoice into a table.
- Translate the document into another language.
- Find it years later in your archive with a single keyword.