Monitoring the Elderly Using Time of Flight Kit

Abstract

It is seen that the physical function of elderly people decreases with age. Most of the problems are related to blood circulation. We need to monitor the state of life before a heart attack or stroke. The falling rate of elderly people also increases. The hustle and bustle of daily existence causes young couples to be bread earners. This causes the elderly people to be left alone without the supervision of others. Activities like frequency of breathing, position inside a room, heart attacks, and stroke are leading causes of mortality in elderly population. Activity monitoring of elderly people provides clues to rehabilitate them before disaster happens. In most cases of long-term attention, tracking of daily activities helps prevent injuries and guarantees safety. In the event of a crisis, late treatment can result in disastrous consequences. Time of flight sensors are used to acquire depth information about the environments. A time-of-flight camera emits modulated laser light and the reflected light from the point to measure distance. Thus, dense depth measurements can be made with low latency. From the time-of-flight data feature, information is extracted and judgmental decisions are made. From the camera, an image frame is obtained and after background subtraction, the target information for extracting the anomaly is obtained. Notable activities which signify important events like wake time, sleep time, and mealtime need to be estimated. An analog device AD-96TOF1-EBZ development kit and a dragon board 410c are employed for the detection and live streaming of depth and IR data is realized. A color map of the depth data with colors going from warm to cold as distance increases is obtained. The depth map is classified using a convolutional neural network.