Image to Text Converter (OCR)

Recognize the text in a scanned page or photo and make it copyable. Supports Turkish and English, processed in your browser.

When should you use this?

Use this when all you have is an image or a scanned page and you need the text inside it to become editable or copyable.

When shouldn't you use this?

If your file is already a text-layer PDF (you can select the text), you don't need this, the PDF-to-Text tool is faster and more accurate. Also, accuracy can drop on very low-resolution images or handwriting.

Step-by-step

Choose the photo or scanned page in the "Image" field.
Pick the text's language under "Language": "Turkish", "English", or "Turkish + English" for documents that mix the two.
Click "Extract text" and wait for recognition to finish.
Copy the text to wherever you need it, then compare numbers and names against the original image.

Example

A student photographs two pages of a library book and wants to quote a passage in an essay. Instead of retyping it, they upload the photo with the language set to English. A few seconds later the paragraphs come back as editable text. Before pasting the quote, they rejoin words that were hyphenated across lines and check that punctuation came through cleanly. If the book mixes English with Turkish terms, the "Turkish + English" option tends to handle that mixed text better.

When is it useful?

It helps whenever text is stuck on paper or inside an image and you want to edit, search or quote it without retyping. Typical jobs include moving details from a receipt into a spreadsheet, copying an address off a poster, or getting an old printed document into a digital archive. Handwriting, decorative fonts and very low-resolution images are where results get unreliable.

Recognising text in an image

The tool turns text in an image into editable text using the Tesseract OCR engine. You can choose Turkish, English or both. JPG, PNG and WebP images are accepted. The image is not uploaded to a server; recognition runs in your browser. However, the OCR engine and language data are downloaded on first use from jsDelivr, a third-party content delivery network, so the first run takes a while and needs an internet connection. The result appears in a read-only text box that you can copy from.

For better results

Clear, straight, well-lit images give the best results. Skewed, blurry or low-resolution images raise the error rate.
Pick the right language. Choose “Turkish” for Turkish text and “Turkish + English” for mixed text.
Handwriting, decorative fonts and tables may not be read reliably. Always compare the result with the original image.
With personal or confidential documents the image does not leave your device, but remember the first download needs a connection. If nothing is recognised the tool says “No text found”.

Things to keep in mind

Shoot the page straight on, in good light, without shadows. Tilted or curled pages can scramble lines.
The wrong language setting can mangle special characters. Turkish letters like ş and ğ need the Turkish option.
Tables and multi-column layouts get read, but the column structure may not survive. Expect some cleanup.
In amounts, account numbers and phone numbers, 0 and O or 1 and l can get mixed up. Always check those by eye.

Frequently asked questions

Is my image uploaded anywhere?

No, your image is not uploaded to a server; processing happens in your browser. Only the model or engine files needed on first use are downloaded from a third-party content delivery network.

Will there be quality loss?

Some quality loss during format conversion or compression is normal, more noticeable with lossy formats like JPG.

Which formats are supported?

JPG, PNG, and WebP formats are supported.