Implementing the solution
You probably noticed that multiple user flows, such as Check Image Quality, contain the same tasks. We can encapsulate that logic into Python functions (or AWS Lambda functions) to avoid code duplication.
Checking image quality
The first step in every user journey is to confirm that the photograph is usable. For our use case, this means the photo contains one person, and they’re looking at the camera.
The DetectFaces API assesses an image and returns the 100 more prominent faces. You’ll receive high confidence scores for frontal faces and notice performance degradations with photos from obscure angles.
Like other Amazon Rekognition APIs, you must specify either a base64-encoded image or a location within an Amazon S3 bucket:
import boto3
region_name = 'us-east-2'
bucket_name = 'ch04-hotel-use2'
image_name = 'images/Nate-Bachmeier.png'
rekognition = boto3.client('rekognition',region_name...