Extract & Verify Book Titles from Bookshelf Photos via GPT-4o & Google Books

This n8n workflow revolutionizes book inventory management by analyzing photos of bookshelves. It starts with a webhook receiving an image URL from a frontend app, saves the image consistently, then uses GPT-4o to detect and extract titles and authors from book spines, even handling partial or obscured text.

Next, it splits the extracted data into individual books, queries the Google Books API via HTTP requests to verify and complete details like full titles, authors, and ISBNs. Results are tidied, aggregated, deduplicated, and returned to the frontend as a clean list. This ensures accuracy beyond basic OCR, reducing manual entry errors.

Benefits include time-saving automation for book collectors, librarians, sellers, or apps like inventory trackers. Use cases: mobile scanning apps (e.g., with Supabase storage), e-commerce book listing, personal library catalogs. Highly scalable for multiple images, with robust error handling via AI and API fallback.