You search a PDF for a word you can see on the page and get no results. Or you try to copy a paragraph and select the whole page as one picture. The PDF is a scan: it contains images of text, not text. Optical character recognition (OCR) fixes that.
What OCR adds to a PDF
OCR software looks at each page image, finds lines and characters, and works out the words. A searchable PDF keeps the original scan exactly as it was and places an invisible layer of recognised text on top, aligned with the words in the image. You see the scan; search, copy and screen readers use the text layer.
Because the image is untouched, nothing about the page’s appearance changes, and a recognition mistake only affects searching and copying, not what you see.
Getting accurate results
- Resolution: 300 dpi is ideal for OCR. Below about 150 dpi, small print starts to fail.
- Straight pages: skewed or curved pages (thick books, phone photos at an angle) reduce accuracy. Flatten the page and shoot straight on.
- Contrast: dark text on a light background works best; avoid shadows across the page.
- Correct language: choose the language of the document. Mixed English and Hindi documents work best with both selected.
Hindi and other scripts
Devanagari is harder for OCR than Latin script: letters join along the top line (shirorekha), and vowel signs sit above, below and beside consonants. Modern neural OCR handles it well on clean printed text. Handwriting and decorative fonts remain unreliable in any language. Our Hindi OCR stores the text in logical order, so copied Hindi pastes correctly into documents and search boxes.
Searchable PDF, text or Word?
- Searchable PDF if you want to keep the document as it looks and be able to find things in it.
- Plain text if you only need the words, for example to paste into an email.
- Word (DOCX) if you need to edit. Expect to fix formatting: OCR recovers words and paragraphs, not exact layouts.
Privacy
Scanned documents are often the sensitive ones: IDs, mark sheets, medical reports. Myriadfile runs OCR in your browser with the open-source Tesseract engine, so pages are never uploaded. The first run downloads the language data, after which recognition happens entirely on your device.
Sources: Tesseract OCR documentation; Unicode Devanagari code chart.