Detection of Fall Events and Rehabilitation Exercises for Elders Using Deep Learning Algorithm – A Brief Survey

Abstract

Gradual increase of elderly population occurs everywhere throughout the world. Fall in elders is a main problem which causes severe injuries and sometimes even to mortality. They can be rescued from this disasters by our immediate attention. The quick detection of fall events in elders prevent them from fall injuries. Computer vision based instruments such as single RGB camera, 3D – based multiple camera, 3D – depth camera can be used to detect fall occurrence in elders. This instruments are very cheap to buy. The use of computer vision methods track human skeleton joints and thereby suitable rehabilitation exercises are given in homely environment. Every year, in western countries more than 2 million of elderly population are given treatment for fall injuries in emergency wards. At present, researchers pay more attention towards finding several systems for fall detection and prevention in elders. In this paper, we are going to survey a brief study of art of fall detection in elderly people and also the use of deep learning algorithms for detection of fall in elderly people.