AI-Powered Image Object Detection and Indexing Workflow

This n8n workflow automates the process of analyzing an image to identify, crop, and index objects within it for advanced image search capabilities. Starting with a manually triggered event, the workflow fetches a source image, then uses Cloudflare’s AI service with the Detr-Resnet-50 model to classify objects within the image. Objects with high confidence (score >= 0.9) are filtered out for further processing. Each detected object is cropped out of the original image to create individual object images, which are then uploaded to Cloudinary for storage. The workflow also creates metadata entries in Elasticsearch that include object labels, URLs, and other relevant information, building a searchable index for object-based image retrieval. This setup is useful for dynamic image cataloging, e-commerce product listing, or digital asset management where detailed image searchability is required.