Face detection example demo

This demo illustrates how to reuse simple pieces from the Ubuntu distribution to detect and count the number of people stopping by at a booth. It has as well a demo-scripts/ directory which contains script to reinstall, update and rollback to show up the rollback mecanism.

Hardware requirement:

You need a webcam attached to your snappy device to try this demo out.

Install and use it:

Install from the store:

sudo snap install face-detection-demo --devmode

Using it:

The service will start automatically and you can head over to http://localhost:8042 to get your camera starts counting persons through face detection.

You will see there a graph updating automatically every 10s with the new data and showing off the last photo where someone was detected.

Settings:

Some settings are available for you to tweak. You need to create a settings file in $SNAP_DATA before the service exits. Note that the reinstall scripts does this if you set the settings file next to it before executing.

The settings file itself is a yaml format configuration, which can contain:

facedetection:
  webserver-port: 8042
  interval-shots: 10

webserver-port enable changing the ports where the web server listen at. interval-shots is the interval between 2 shots in seconds.