Image object classification and indexing in Elasticsearch

• Automate image processing workflows for e-commerce product images.
• Enhance search capabilities by indexing images based on detected objects.
• Streamline content management by organizing images in a database.

Set up initial variables including source image and model. Fetch the source image from a specified URL. Classify the image using the Detr-Resnet-50 model to identify objects. Filter results based on a confidence score to ensure accuracy. Crop identified objects from the original image. Upload cropped images to Cloudinary for storage. Index the uploaded images in Elasticsearch for improved search functionality.

1. 1
   Import the workflow template into your n8n environment.
2. 2
   Set your variables in the 'Set Variables' node, including the source image URL.
3. 3
   Execute the workflow by clicking 'Test workflow' to start the image processing.