Does the original image information in the PDF need to be parsed?

Filimoa / open-parse

Improved file parsing for LLM’s

https://filimoa.github.io/open-parse/

MIT License

2.18k stars 83 forks source link

Does the original image information in the PDF need to be parsed? #37

Open ic-xu opened 2 months ago

ic-xu commented 2 months ago

Description

PDF is a document with mixed graphics and text. When we are doing RAG, the pictures in the PDF often contain important information, so we generally need to return the parsed pictures to the user as they are; currently I have a private version that makes pictures Extraction, I am not sure whether the main branch needs this part of the function

Filimoa commented 2 months ago

Open a PR and we can take a look!