Translate Image reads the printed text in a photo or screenshot with OCR in your browser, then sends that extracted text to Google Translate and shows the translation. The OCR recognises English and Bangla; the result is plain text, not an edited picture.
It works in two stages. First, Tesseract OCR runs inside your browser and turns the letters in the image into text. Second, only that text is sent to Google Translate, which detects the source language and returns the version you chose: Bangla, English, Arabic, Hindi, Urdu, Spanish, French, German, Chinese (Simplified), Japanese, Korean, Russian, Portuguese, Turkish or Italian.
The OCR step is tuned for English and Bangla, which covers notices, product labels, forms and screenshots most people in Bangladesh deal with. Text in other scripts, such as Arabic or Chinese, will not be read well from the picture, even though those languages are available as translation targets.
Add the picture
Choose a photo or screenshot that shows printed English or Bangla text.
Pick a target language
Select the language you want from the list; Bangla is selected by default.
Run it
OCR runs first (the progress bar shows each stage), then the text is translated.
Copy or download
Copy the translated text or save it as a .txt file with the language code in its name.
ছবি বা স্ক্রিনশটে থাকা ইংরেজি বা বাংলা লেখা পড়ে সেটাকে আপনার পছন্দের ভাষায় অনুবাদ করে দেয় এই টুলটি। প্রথমে আপনার ব্রাউজারেই OCR দিয়ে লেখা বের করা হয়, তারপর শুধু সেই লেখাটুকু Google Translate-এ পাঠিয়ে অনুবাদ আনা হয়। ইংরেজি নোটিশ, ওষুধের কাগজ বা অফিসের চিঠি বাংলায় বুঝে নিতে কাজে লাগে। ফলাফল আসে টেক্সট হিসেবে, ছবির ওপর লেখা বদলায় না।
হ্যাঁ, ভাষা হিসেবে বাংলা বেছে নিলে অনুবাদটা বাংলা টেক্সটে আসবে। মেশিন অনুবাদ, তাই জরুরি কাগজে একবার মিলিয়ে নিন।
ছবি ব্রাউজারেই থাকে, কিন্তু ছবি থেকে পাওয়া লেখাটা অনুবাদের জন্য Google Translate-এ যায়।
The image stays in your browser, where the OCR runs. The text read from it is then sent to Google Translate to be translated, so do not use this for text you would not want Google to process.
No. The tool returns the translation as plain text that you can copy or download. It does not paint new words onto the picture.
Not reliably. The OCR stage recognises English and Bangla only. Arabic, Chinese and other languages are available as translation targets, but text written in those scripts will not be read correctly from an image.
Two machines are involved: OCR can misread a letter, and Google Translate then works from that text. Check the extracted words when the result looks odd, and use a clearer, straighter photo where you can.
The first time, the OCR engine and language data (about 10 MB) download and are then cached. After that, reading a typical photo takes 20 to 60 seconds, and the translation itself takes a second or two.
Yes, there is no charge or sign-up.
More like this in Image tools. Need a version built for your business? Ask us.