Show HN: OCR Workbench: AI OCR for hard documents
github.comOCR on old documents is hard. OCR Workbench uses AI for OCR and provides an editing environment to clean things up, as is inevitably required.
Inspired by this Hacker News post: https://news.ycombinator.com/item?id=43048698
Backstory: I was having trouble producing transcriptions of Colonial American documents, which have their own unique challenges for OCR, and things like Tesseract fail miserably. So I built something. Uses Gemini and seems to work pretty well (disclaimer: you need your own API key). I didn't build Claude but I expect it works similarly well.
FWIW: largely vibe coded, with human review and intervention as required. Interesting concept. I tried it with a text written in Church Slavonic, didn't work. I guess the documents don't have to be THAT old. It would also be nice if you could upload images individually instead of selecting everything from a folder. Either way, nice work. Thanks! I have not tried anything other than English, not sure how good the LLMs are for that. Did you use Tesseract or Gemini? Once the page structure is set up from the images (via the directory upload), you can upload new images for each page, but I didn't include an option to just create all the pages manually. It's a good idea. Going to add that... I updated the app to allow manual add/delete of pages. Then tap the page and do New Image to upload per page. Thanks for the feedback!