Published on

Using Python for Image Analysis with Amazons Rekognition Service


Amazon's Rekognition Service, which falls under their Artificial Intelligence tier, makes it easy to add image analysis to your applications.

Today we will use Rekognition to analyze an image, to determine the percentage of detection that the service analyzes. We will be using the Python SDK to do this.

Getting a Random Image:

So, I got this drunk guy on the couch, which I thought we could use to analyze.

Image Used:

Our Python Code:

Our code will use boto3 to use rekognition from Amazon Web Services, detects the image, and prints out the values.

Note that I am not specifying any credentials, as my credentials is configured in my local credential provider, where boto will pick it up from.
import boto3

BUCKET = "rekognition-bucket"
KEY = "images/image-02.jpg"

def detect_labels(bucket, key, max_labels=10, min_confidence=90, region="eu-west-1", profile_name="aws"):
    rekognition = boto3.client("rekognition")
    response = rekognition.detect_labels(
        "S3Object": {
        "Bucket": BUCKET,
        "Name": KEY,
    return response['Labels']
for label in detect_labels(BUCKET, KEY):
    print("{Name} - {Confidence}%".format(**label))

Running the App:

Running our Python App, will result in the following:

$ python 
People - 98.9893875122%
Person - 98.9893951416%
Human - 98.9505844116%
Alcohol - 98.573425293%
Beer - 98.573425293%
Beer Bottle - 98.573425293%
Beverage - 98.573425293%
Bottle - 98.573425293%
Drink - 98.573425293%
Couch - 98.4713821411%


Thank You

Thanks for reading, feel free to check out my website, feel free to subscribe to my newsletter or follow me at @ruanbekker on Twitter.

Buy Me A Coffee