Lab#4
Lab#4
Rubric
Functionality Marks
Lambda function to detect labels, and write the labels and other corresponding 5
image metadata to DynamoDB table
1. Create DynamoDB table via AWS Management console (1 mark)
2. Data model for input (1 mark)
3. Get newly uploaded image(s) object URL (1 mark)
4. Insert an item for each uploaded image object url and all detected labels
with confidence level that is higher than the specified level to the
DynamoDB table, use map data type to store label name and confidence
(2 marks)
Lambda function to generate thumbnail and store it to S3 5
1. Generate thumbnail (3 marks)
2. Save the thumbnail (2 marks) to S3
State machine 5
Publish to AWS Lambda 2
Automate your state machine (finish the tutorial) 5
The screenshot to show that you have deleted your AWS account 3
Question
You are asked to develop a serverless application using AWS step function to build an image processing
workflow. More specifically, your serverless application should be triggered when image file(s) have
been uploaded to S3 bucket, and do following:
1. Use AWS Rekognition service to detect label and insert those labels with confidence greater
than specified level to your DynamoDB table
Lab #4 Page 1 of 2
API Engineering COMP306
2. Generate a thumbnail. In order to avoid recursive invocation, store the generated thumbnail in a
different S3 bucket. you can find code snippet to generate thumbnail at
https://www.grapecity.com/blogs/create-a-thumbnail-image-using-documents-for-imaging
Lab #4 Page 2 of 2