What is Redactopix?
Redactopix reads text inside images with OCR and turns it into editable text. You can copy or edit the recognized text and save it as a .txt file. Drag a box on the preview to recognize just that area for partial reading. To improve scan accuracy you can switch the noise-filter strength (off, mild, normal, strong). It recognizes Japanese and English at the same time, powered by the tesseract.js engine.
How to use
- Pick the image whose text you want to read, or drag and drop it in.
- On first use the OCR engine and language data (Japanese and English) load from a CDN. After that they start from the cache.
- Choose whether to read the whole image, or drag a box on the preview to limit the recognized area.
- If the text is faint or the background is noisy, switch the noise-filter strength (off, mild, normal, strong).
- Run OCR and the recognized text appears in the results panel.
- If there are misreads, edit them directly in the results panel to clean up the text.
- Copy the text to paste into another app, or save it as a .txt file.
Great for
Transcribing business cards
Read the company, name, and contact details from a photographed business card and copy them into your contacts.
Digitizing documents
Photograph paper materials or handouts and turn them into text you can search and reuse later.
Recording signs and notices
Capture the wording of a storefront notice or information board as text to reread or summarize into notes.
Extracting text from screenshots
Pull just the text you need out of an app screen or screenshot where copying is not allowed.
Entering receipts
Read the amount and store name from a receipt to cut down on manual entry for budgeting or expense reports.
Reading only part of an image
Box in a specific paragraph or heading on a page so you can transcribe just that, excluding unwanted areas accurately.
Privacy
Images are analyzed inside your browser with WebAssembly and are never sent to a server. Only on first use are the OCR engine and language data (Japanese and English) fetched from a CDN; that is a static file download, not your image or recognition results being sent anywhere.
FAQ
- Q. Are my images uploaded?
- A. No. OCR runs inside your browser with WebAssembly. Neither the image nor the recognition results are sent externally; everything is processed on your device.
- Q. Which languages are supported?
- A. It supports Japanese and English at the same time. Even if a single image mixes Japanese and English, it can recognize them together.
- Q. Can I read only part of an image?
- A. Yes. Drag a box on the preview to OCR just that area, letting you transcribe only the paragraph you need accurately.
- Q. What is the noise filter for?
- A. It suppresses background grime, shadows, and faintness to improve accuracy. Choose from four levels, off, mild, normal, or strong, to match the image.
- Q. What if the recognition is wrong?
- A. You can edit the results directly in the panel. Fix any misread characters before copying or saving as .txt.
- Q. Can it read handwriting?
- A. The engine is built for printed type, so handwriting accuracy drops sharply. Clean printed text gives the most stable results.
- Q. How can I save the results?
- A. You can copy the text to paste into another app, or export and save it as a .txt file.
- Q. Is it free?
- A. It is free to use. No registration is required, and you can run OCR, partial reading, and text export without a usage limit.