adding the Amazon Rekognition component

In this exercise, you will extend the application by adding the Amazon Rekognition component. As soon as you upload a photo to your Amazon S3 bucket, Amazon Rekognition processes the photo and identifies objects, people, text, scenes, and activities in the photo and labels it accordingly. 
Note: Make sure to sign in to your AWS account with the AWS IAM user edXProjectUser credentials.

To get started, follow the instructions below.

1. Download the exercise code .zip file to your AWS Cloud9 environment.

• Type the command below in your AWS Cloud9 terminal to make sure you are in the ~/environment directory of your AWS Cloud9 instance.
  cd ~/environment
• Type the command below in your AWS Cloud9 terminal to download the python code for the Amazon Rekognition component.

wget https://s3-us-west-2.amazonaws.com/us-west-2-tcdev/courses/AWS-100-ADG/v1.0.0/exercises/ex-rekognition.zip

On the left tree view, you should see the exercise code .zip file named ex-rekognition.zip.

2. Unzip the exercise code .zip file.

• Unzip the exercise code .zip file by typing the command below on your AWS Cloud9 terminal.

unzip ex-rekognition.zip

The contents of the .zip file should be extracted to a folder with a similar name. You can view the folder on the left tree view.

• You may want to close any tabs that remain open from previous exercises.

3. Explore the exercise code.

• Open the exercise-rekognition/FlaskApp/application.py file.
• In the Homepage route function, notice that a Boto 3 client for Amazon Rekognition is created. The image uploaded in the Amazon S3 bucket is passed to the detect_labels API, which returns a list of labels processed by Amazon Rekognition. These labels are then populated on the UI.

4. Run and test the code.

• To run the application.py code, on the top menu bar, click Run -> Run Configurations -> Python3RunConfiguration.
• Important: Notice that the run configuration runs the application.py for the previous exercise.
• Point the run configuration to the correct exercise folder by editing the folder path in the Command text box in the bottom pane.
  In that text box, type exercise-rekognition/FlaskApp/application.py
• Click Run on the left side. You should see a message like this:

Running on http://0.0.0.0:5000/

• Go to your browser and type the IP address of the Amazon EC2 instance that hosts your AWS Cloud9 environment. At the end of the IP address, type :5000

  The application should now have the functionality related to Amazon Rekognition.

• To test the Amazon Rekognition component, click Home on the application.
• Upload an image. Amazon Rekognition should label the image with the image properties.

Optional Challenge

The Boto 3 detect_labels response includes a Confidence value. Can you update the application UI to include the Confidence? Or define a threshold and only display labels over the confidence threshold?

眼鏡蛇

posted on 2018-04-19 11:16 眼鏡蛇 閱讀(161) 評論(0)  編輯  收藏 所屬分類: AWS