Step 7: Testing the application

App Status

You now have an application listening at the route that was created during the deployment. You can test it but simply clicking on the route link, or copy/paste it in your browser:

alt text

Uploading images

As our application is now a REST API endpoint, there are multiple ways to upload images to it. Here are a few. Note: as our prediction function is waiting for a json payload with a base64 encoded image, we must do this encoding first.

CURL on Linux or Mac with bash/zsh

From anywhere you have an example image like car.jpg (replace with the right name in the command, as well as the Route with /predictions at the end):

(echo -n '{"image": "'; base64 car.jpg; echo '"}') | curl -H "Content-Type: application/json" -d @- http://licence-plate-workshop-git-lpr-workshop.apps.rhods-test.rqdu.p1.openshiftapps.com/predictions

Invoke-WebRequest on Windows with Powershell

From anywhere you have an example image like car.jpg (replace with the complete path and name in the command, as well as the Route with /predictions at the end):

Write-Output ('{"image": "' + ([Convert]::ToBase64String([IO.File]::ReadAllBytes('C:\Users\Guillaume\Downloads\car.jpg'))) + '"}') | iwr -Uri http://licence-plate-workshop-git-lpr-workshop.apps.rhods-test.rqdu.p1.openshiftapps.com/predictions -Method 'POST' -ContentType: 'application/json' | Select-Object -Expand Content

From a notebook

Open the notebook named 05_Send_image.ipynb
Upload an image in your environment
In the first cell, replace the placeholders with your:
- Image path and name.
- Route to the service.
Run the cells and see the result!

alt text

Once you’re finished, you can come back here and head to the next section.

Red Hat OpenShift Data Science Workshop

Licence plate recognition workshop