ShelfCheck: a bookshop shelf photo that ends in a useful checklist

Description and use

ShelfCheck enables independent bookstores to quickly audit shelf placement using browser‑based OCR and generate a checklist of confirmed moves and possible gaps.

Problem

Staff need to locate misplaced books and missing facings on a shelf without a full inventory scan.

Outcome

ShelfCheck turns a single shelf photo and catalog CSV into a concise staff checklist of confirmed misplacements and potential gaps, enabling quick shelf audits without server‑side processing.

Workflow

• Upload catalog CSV and shelf photo
• Draw a box around a book spine
• Run OCR on the cropped spine at 0°, 90°, and –90°
• Rank OCR output against catalog titles/authors using character‑trigram similarity
• Confirm the suggested match
• Generate a Move task if the book is in the wrong section
• Repeat for all visible spines
• Mark review complete to identify Possible gap tasks for missing facings

Evidence

ShelfCheck ranks catalog titles, then asks a person to confirm the match.

Source

The app never calls a photo proof of a stockout.

Source

Tesseract.js is the open‑source AI core. It runs an OCR model in the browser; no server receives the photo or catalog.

Source