Fall detection using pose estimation models and smart alarming system

The number of aged people is constantly increasing in India and other countries widely as India saw a rise in the older population from 8.6% in 2011 to 10.1% in 2021 and it will be 13.1% in 2031. This can infer that the older population will increase rapidly. There should be a model which detects falls of these old aged persons because at any moment they can fall. In this proposed model there are 3 phases which are pose estimation, fall detection, and alarming system. In pose estimation, this proposed model used OpenPose, PoseNet and MoveNet. Fall detection has been identified by an abrupt change in the position of a person’s coordinates calculated by pose estimation models and for the alarming system proposed model used a WhatsApp chatbot and mobile phone calling for alarming the corresponding person.