In this demonstrator, the EoT device can work in one of two modes. In the ‘normal’ mode, the camera captures a snapshot at regular intervals (configurable by the user). Snapshots are stored on the SD card. In the ‘smart’ mode, the camera performs an analysis of the image, trying to detect ‘key events’ that are interesting to the user. Regions of interest can be also drawn to detect motion. Whenever one or more of the key events is detected, the image is stored in the SD card, along with text information (tags) indicating the events recognized. There is also the option to send push notifications to a smartphone. An Android App allows to retrieve the images and their associated tags, and upload them to a cloud database. The cloud database can be accessed from any standard browser. Here the user has additional functionality, presented in a Zapier/IFTTT manner to allow further processing or distribution of the images. The images below show some sample screenshots of the Android App, a user drawing a region of interest and access to captured images from a web browser.
Litterbug is an incarnation of the Versatile Mobile Camera that aims at detecting illegal littering on streets.