Collect, AI-Label Images & Store in Google Sheets

This n8n workflow automates the collection of images from Google Custom Search based on a search query (e.g., 'street'), analyzes them with AWS Rekognition to detect and extract labels, formats the data (image name, link, and labels), and appends it as rows to a specified Google Sheet. It's ideal for rapidly creating labeled image datasets without manual effort.

The process starts with an HTTP Request node querying the Google Custom Search API for photo-type images. The binary image data is then passed to AWS Rekognition's 'detectLabels' operation, which identifies objects, scenes, or concepts in the images. A Set node extracts and structures key fields like image title (as name), link, and label names into a clean JSON object. Finally, the Google Sheets node appends this data to your sheet using OAuth2 authentication.

Benefits include massive time savings in data preparation for machine learning projects, market research, or content analysis—handling search, AI labeling, and storage in one triggerable flow. Use cases: Building training datasets for computer vision models, monitoring visual trends (e.g., urban scenes), HR photo cataloging with auto-tags, or e-commerce product image research. Customize by changing the search query, sheet ID, or adding filters for specific labels.

With only 4 nodes but powerful integrations (Google API, AWS AI, Google Sheets), it's efficient, scalable, and requires minimal setup after credential configuration. Perfect for non-coders needing pro-level vision automation.