Automate Image Labeling with Amazon Rekognition


112 students completed the lab in ~36m

Total available time: 1h:0m

40+ students rated this lab!

How to automate object detection with Amazon Rekognition

Amazon Rekognition allows you to detect objects and scene details from images. It provides a stateless and secure API that simply returns a list of related labels, with a certain confidence level.

In order to automate the labels extraction process, we will build a serverless system that runs object detection on every image uploaded to S3. We will use AWS Lambda for the processing logic, and save the results on a DynamoDB table.

We will see how to setup S3, DynamoDB and AWS Lambda to automate the whole process. During this laboratory, we will also discuss its API details, optional parameters, best practices, etc.

Follow these steps to learn by building helpful cloud resources

Log In to the Amazon Web Services Console

Your first step to start the Lab experience

Object Detection Context and Limitations

What is Object Detection and how does it work?

Create an S3 bucket

Create a new S3 bucket

Create a new Lambda Function with an S3 trigger

Connect a Lambda Function to the source S3 Bucket.

Implement the Object Detection logic

AWS Lambda implementation of object detection with Amazon Rekognition.

Test the labeling system with new images

Verify that the system is working properly by uploading new images.