Image Processing
Deploy an AI/ML Image Processing Pipeline
Build, deploy, and operate a high-speed image tagging and processing solution using pre-trained AI models and serverless functions
Who this is for
What you'll deploy
How you'll deploy
What is image processing?
Image processing uses software to identify, classify, tag, transform, or modify the contents of images at scale. It can include object recognition and detection, grouping similar images together via classification, adding searchable metadata, or bulk resizing or sharpening
What does image tagging (annotation) do?
Image annotation uses pre-trained machine learning models to identify objects within images. Images are tagged with metadata that can then be used to sort, filter, classify, segment, group or otherwise more intelligently use the images for common needs.
What are common use cases for high-speed image processing?
Simple image processing can prepare images to load faster in mobile apps or websites, and AI-driven image processing can generate insights from images to detect defects in manufacturing, remove unsafe imagery from user-generated content, or diagnose diseases in healthcare.
Deploy event-driven image processing using Cloud Functions and Google Cloud Vision
Learn how to use an event-based architecture to take uploaded files and process them using Cloud Vision
- User calls the REST API with image reference and optional Cloud Vision annotation features.
- This triggers a Cloud Function, which calls Cloud Vision to detect and annotate the image. REST API returns annotations (JSON) to the user.
- User batch uploads images to Cloud Storage.
- The upload event triggers a Cloud Function.
- Cloud Function calls Cloud Vision to detect and annotate the image.
- Cloud Function stores the resulting prediction and annotation for the image in Cloud Storage.
Google Cloud Experience Level
Beginner
Estimated deployment time
12 min
2 min to configure, 10 min to deploy
- Active Google Cloud account
- Administrator rights to your project
