Umi-OCR setup: screenshots, batch text and QR codes

This Chinese-language tutorial, uploaded on January 4, 2024, demonstrates Umi-OCR for extracting text from screenshots and saved images on Windows. The speaker describes it as free, open source software with offline batch recognition. No release number or recording date is supplied, so the download details describe the release shown in this upload.

The setup walkthrough compares two downloads: a 118 MB package that the speaker says does not support Pentium, Celeron or Atom CPUs, and an 86.7 MB package described as using less memory but running more slowly. The downloaded executable extracts into a folder, where the presenter opens the main program.

The screenshot workflow starts with selecting a screen region. Umi-OCR displays recognized text alongside the image, and the presenter edits the output before copying it. The tutorial also demonstrates pasting images, assigning a screenshot keyboard shortcut and dragging local image files into the app. Handwriting tests produce mixed results; one sample needs corrections. These examples do not establish a general accuracy rate.

For batch OCR, the presenter selects images or a folder and starts a recognition task. A QR-code example returns a link, while a barcode example returns digits rather than product information. The speaker also explains switching from Tianruo OCR after its free edition stopped working for them; this is a dated personal report.