Documentation

If the Android device running Slideshow app has a camera, it can be used for detecting how many people are currently in the area in front of the camera. This number can be used for estimating how many people can see what is on the screen at the moment and for either triggering a content or for further analytical and statistical purposes.

If you would like to further integrate the face detection feature with an external system, please contact us.

Setup

Following steps are needed for enabling face detection:

  1. On-screen menu – Basic settings – Request camera permissions – allow Slideshow to access the camera on the Android device. If this option is missing from the Basic settings, the permission has been already granted and it is not necessary to grant it again.
  2. Web interface – menu Settings – Device settings – select input camera under Camera for face detection and setup other options if needed.
  3. Reload Slideshow app in order to apply the settings.
  4. Wait 10 seconds after startup, navigate to web interface – menu Information – Face detection to check whether the camera image is correct. Refresh this page in order to get updated image and statistics.

Usage

The quick overview of the face detection status can be seen on the Home page of the web interface. If the face detection is active, Face detection widget containing statistics will be displayed. By clicking on the Number of faces link detected in the last frame (alternatively through menu Information – Face detection) you can open a page with detailed statistics for the last 60 frames, as well as the last processed image from the camera. All detected faces together with their tracking IDs are marked in red and probability whether each eye is opened is marked with blue and green.

Results of face detection can be used in Triggers to trigger a change in playback. For example detecting a person close to the camera can trigger a different playlist on the screen.

Face detection can’t be used at the same time as video input from the same camera, as two different processes can’t use the same camera at the same time.