Amazon rekognition service usage via the aws cli

rekognition.md

Rekognition

Amazon announced their Rekognition image analysis service in Nov-2016. The documentation page at that time did not include a specific example of how to use it via the aws cli so I thought that it might be helpful to show an example.

The aws cli must already be setup and usable.

The picture is of my dog laying down and waiting for his dinner.

Replace bucketname and imagename with an s3 bucket and image file that are accessible to the cli.

Usage:

aws rekognition detect-labels --image='{"S3Object": { "Bucket": "bucketname", "Name": "imagename" } }'

Response

{
"Labels": [
    {
        "Confidence": 64.35530090332031,
        "Name": "Animal"
    },
    {
        "Confidence": 64.35530090332031,
        "Name": "Canine"
    },
    {
        "Confidence": 64.35530090332031,
        "Name": "Dog"
    },
    {
        "Confidence": 64.35530090332031,
        "Name": "Mammal"
    },
    {
        "Confidence": 64.35530090332031,
        "Name": "Pet"
    },
    {
        "Confidence": 64.35530090332031,
        "Name": "Pointer"
    },
    {
        "Confidence": 62.247352600097656,
        "Name": "Beagle"
    },
    {
        "Confidence": 62.247352600097656,
        "Name": "Hound"
    }
]
}