OCR Explained: Turn Scans and Photos Into Editable Text
Published
OCR stands for optical character recognition, and it solves a specific, narrow problem: turning an image of text into text a computer can actually read. A page looks like text to your eyes either way — printed or handwritten, scanned or freshly typed — but to a computer, a scanned page and a typed document are entirely different things until OCR bridges that gap. Knowing when you need it, and what kind of result to expect, saves a lot of frustration with files that look right but behave strangely: a scanned PDF that won’t let you search or copy, a photographed page you can’t paste into a document.
What OCR actually does
To a computer, an image is just a grid of colored pixels — it has no idea some of those pixels happen to form the letter “A” unless something tells it so. OCR software analyzes the shapes in that grid, matches them against patterns it has learned for letters and numbers, and outputs the text it recognizes. Depending on the tool, that recognized text either comes back on its own, as a plain file you can copy and edit, or gets layered invisibly back on top of the original image, so the image still looks the same but now has real, selectable text hiding underneath it.
When you actually need it
A quick way to tell whether a document needs OCR: try selecting a line of its text. If your cursor highlights individual words the way it would in a text editor, the document already has real text, and OCR won’t add anything useful — it’s redundant on a file that’s digital already. If dragging your cursor over the words does nothing, or selects the whole page as one block, you’re looking at an image standing in for text, whether that’s a scanned form, a PDF exported from a photocopier, or a phone photo of a printed page. That’s the situation OCR is built for.
Pulling text out of a single photo, step by step
Image to Text reads a single photo, screenshot, or scan and hands back the recognized text as a plain file, using a tesseract-based recognition engine on Adawaty’s server.
- Open Image to Text and choose one photo, screenshot, or scan.
- Pick the language — English, Arabic, or English + Arabic for a mixed document.
- Click Extract text and wait a few seconds while it processes.
- Download the result as a .txt file, ready to paste into any document or note.
Each image can be up to 15 MB, and both your upload and the extracted text are wiped from the server automatically within about an hour of finishing.
Making a whole PDF searchable instead
Image to Text and OCR PDF solve two different versions of the same underlying problem. Image to Text pulls the recognized words out of a single photo and hands them back as a separate text file — the original image doesn’t change, and what you receive is the words on their own. OCR PDF, by contrast, works on an entire PDF and leaves it as a PDF: it recognizes the text on every scanned page and adds it back in as an invisible layer, so the document still looks exactly like the scan it came from, but every page can now be searched, selected, and copied. Reach for Image to Text when you just need the words out of one image; reach for OCR PDF when the whole document has to stay a PDF but start behaving like one.
What accuracy to actually expect
OCR does well on what it was built for: clean, printed text, photographed or scanned straight-on under even lighting. Expect close to perfect recognition on a typed page, a printed textbook, or a clearly lit form. It does noticeably worse on handwriting, which varies too much from person to person for the recognition patterns to match reliably, and on stylized or decorative fonts, which don’t resemble the standard letterforms the engine was trained on. Arabic recognition in OCR PDF is currently in beta and works best on clean, printed Arabic, so treat the output as a draft to proofread rather than a finished transcript, especially on anything handwritten or low quality.
FAQ
Can OCR read handwriting reliably?
Not reliably — handwriting varies too much from person to person for pattern-based recognition to hold up consistently. It occasionally does fine with very neat, printed-style handwriting, but treat any handwritten result as a rough draft to check by eye rather than a finished transcription.
Does running OCR change how my document looks?
No, the original image or scan stays exactly as it was in both tools. OCR PDF adds its recognized text as an invisible layer on top of the existing page images, and Image to Text leaves your photo untouched, returning the recognized words separately as their own file.
What happens to my file after it’s processed?
Both tools run on Adawaty’s server and remove the upload, along with anything extracted from it, on an automatic timer of about an hour. No one reads through the contents in the meantime.
Takeaway
OCR turns an image of text into text a computer can actually work with, and the right tool depends on what you want back: Image to Text for the words out of a single photo, OCR PDF for a whole scanned document that needs to stay a PDF but become searchable. Expect strong results on clean, printed text and weaker ones on handwriting, and proofread anything that matters before relying on it.