官术网_书友最值得收藏!

Invoking the Rekognition service using the AWS CLI

Now, let's invoke the object detection capability of Amazon Rekognition via the AWS CLI. This time, let's perform the object detection on an image that's stored in one of our S3 buckets. We will be using a sample image from Pexels, a website with thousands of royalty-free images. Download the image at https://www.pexels.com/photo/animal-beagle-canine-close-up-460823/ and upload it to the contents S3 bucket.

Here, we can see an adorable beagle puppy laying on what appears to be a bed of gravel: 

You should see the following output when you list the objects in your contents bucket:

$ aws s3 ls s3://<YOUR BUCKET>
2018-12-02 13:31:32 362844 animal-beagle-canine-460823.jpg

Now that we have an image, we can invoke the object detection capability of Rekognition via the following CLI command. Note that we must escape the { and } characters with a \, and we must not include any spaces when specifying the S3 object on the command line:

$ aws rekognition detect-labels --image S3Object=\{Bucket=<YOUR BUCKET>,Name=animal-beagle-canine-460823.jpg\}

The results come back almost instantly:

{
"Labels": [
{
"Name": "Mammal",
"Confidence": 98.9777603149414
},
{
"Name": "Pet",
"Confidence": 98.9777603149414
},
{
"Name": "Hound",
"Confidence": 98.9777603149414
},
{
"Name": "Dog",
"Confidence": 98.9777603149414
},
{
"Name": "Canine",
"Confidence": 98.9777603149414
},
{
"Name": "Animal",
"Confidence": 98.9777603149414
},
{
"Name": "Beagle",
"Confidence": 98.0347900390625
},
{
"Name": "Road",
"Confidence": 82.47952270507812
},
{
"Name": "Gravel",
"Confidence": 74.52912902832031
},
{
"Name": "Dirt Road",
"Confidence": 74.52912902832031
}
]
}

The output is in JSON format, just like we configured the CLI to the output. From the output, the Rekognition service detected several objects or labels. Rekognition is very sure that it detected a dog; it even identified the breed of the dog as a beagle! Rekognition also detected the gravel in the image, which could be a part of a dirt road. The AWS CLI can be very useful when trying out an AWS service, and to see how the output is structured when we are developing our applications.

主站蜘蛛池模板: 饶平县| 东丰县| 依兰县| 清流县| 黄梅县| 鸡泽县| 德钦县| 兰州市| 襄城县| 萍乡市| 赤壁市| 淮南市| 巴彦淖尔市| 锦屏县| 尤溪县| 宽甸| 无为县| 大方县| 神池县| 建始县| 吉首市| 麦盖提县| 普宁市| 渭源县| 绩溪县| 鲁山县| 嘉义县| 奉新县| 奉节县| 乌鲁木齐市| 沛县| 孟津县| 金阳县| 安岳县| 水富县| 郸城县| 鄯善县| 崇文区| 土默特左旗| 全椒县| 浑源县|