Product Image Analysis

AI-powered assessment of product image quality across the catalog

Analysis Settings

Batch Size

Products per run (max 20 — vision AI is slower)

Skip Rows (Resume)

Offset to resume from

Only analyze products not yet assessedAutomatically remove low-quality / broken images (clears image_url)

How it works: Each product image is sent to a vision AI model which rates its quality and flags issues like watermarks, blurriness, or cluttered backgrounds.

Results saved: Each analyzed product gets image_quality_score and image_needs_replacement fields saved to the database.

Tip: Run small batches (5–10) first to calibrate, then increase. Once you have barcodelookup.com API access, we can automatically replace flagged images.