QUIET COPY GUIDE
How does on-device OCR handle your page?
After you add a page or image reference, Quiet Copy reads it in the browser, finds likely text regions and can optionally use local OCR to label them.
01
Projection and adaptive-threshold splitting happens first and does not require OCR.
02
When OCR is enabled, the browser loads Tesseract.js and language data shipped with the site.
03
Recognition results and cropped tiles may be stored in local IndexedDB. The current version has no server-side cloud backup.