Fall detection system for older adults

Abstract

Fall is a major health issue among older adults. Fall may cause serious injury to the older adults such as heap fractures and head traumas. A method for detecting falls among older adults is presented which uses the Microsoft Kinect sensor. The system consists of two parts. The first one is to detect fall using Kinect sensor and second part is to generate control signal to send message. To detect fall condition, motion related features are calculated. For this sensor's color and depth output is used. Ground event segmentation is done to detect moving objects from the video stream followed by feature extraction and event classification. After fall detection, help message will send to the registered number so that immediate medical assistance will get. The system works on real time data.