Umi-OCR — офлайн-распознавание текста с экрана и файлов

★ 46.7K

Umi-OCR is an open-source text-recognition program that works offline on Windows and Linux. Reach for it for quick, everyday recognition: select a screen area with a hotkey and get the text, recognise a batch of images at once, pull text out of a PDF, capture text from a screenshot or a photo. Everything is computed locally, with no internet and no uploading of pictures to someone else's server — suitable for work documents too. The difference from recognition libraries is that this is a finished program with an interface: download it and use it, with no code or pipeline setup. It supports table recognition and many languages. It is handy for anyone who regularly lifts text off pictures and the screen — from a PDF with no text layer, from scans, or from someone else's slides.