The video contains 4 visualization areas, the top row shows a simple context in which a person enters in the scene and falls. On the left there is a sequence of depth frames provided by the Kinect sensor: the algorithm recognizes the human and tracks him with a blue circle (red when a fall occurs). On the right the trajectory of the person’s central point is sketched. The bottom row shows a more complex fall situation, where the person interacts with objects (desk and chair).
