Custom Applications

• Run either the tensorflow image classification or object detection example on the RPi. This requires access to a raspberry pi camera.

• Modify the tensorflow image classification or object detection example to work with images taken from a standard webcam. This will require changing the use to use OpenCV for the image samples rather than the picamera python module.

• Define a custom application that will analyze some data on the edge device and send events (alerts, notifications, etc.) to an application running in the cloud. The cloud application should visualize the events. Verify that the entire processing chain works correctly.