Hasty Briefsbeta

ScribeOCR – Web interface for recognizing text, OCR, & creating digitized docs

a day ago
https://github.com/scribeocr/scribeocr
Copy Link
  • #digitization
  • #OCR
  • #proofreading
  • Scribe OCR is a free web application for recognizing text from images and creating fully-digitized documents.
  • Primary use cases include adding searchable text to PDFs, proofreading OCR data, and creating ebook-style PDFs.
  • Unlike other tools, Scribe OCR allows easy correction of OCR errors and precise text positioning over images.
  • Scribe OCR can be run via the public site scribeocr.com or locally using npm and a local HTTP server.
  • The application focuses on efficient proofreading, improving OCR accuracy from 98% to 100%.
  • Scribe OCR generates custom fonts for each document to improve alignment and error visibility.
  • Three versions of a scanned book page demonstrate Scribe OCR's capabilities: original image, proofreading mode, and ebook mode.
  • Ebook Mode produces small file sizes while faithfully representing the original document.
  • Documentation is available at docs.scribeocr.com, and user feedback is encouraged via Git Issues.
AboutLogin