ClearFrame’s large-batch milestone made the browser-local cleanup workflow practical beyond a handful of photos, while keeping every automatic detection open to human review.
Detection with honest limits
OCR and image geometry look for URL and watermark text, including text placed in rotated corners. Detected regions are shown directly on the image so a person can confirm what the browser found.
Missed regions can be marked manually. Blur, pixelate, cover, or crop can then be chosen for the reviewed region while retaining original dimensions where possible.
Many formats and many files
JPG, PNG, HEIC, and HEIF photos can enter the same batch. All image processing remains in the browser rather than uploading the batch to a cleanup server.
Resilient exports
For very large jobs, ClearFrame can stream ZIP parts incrementally into a chosen folder, using groups of 500 photos and writing a completion note when finished.
Where direct folder writing is unavailable, numbered ZIP parts are downloaded instead. Completed parts remain usable if a later stage is interrupted; a 3,001-entry archive test completed without reported compression errors.