Survey of fall detection and daily activity monitoring techniques

The risk of sustaining heavy injuries through accidental falls creates a major medical problem for elderly people. This paper conducts a survey of the various automatic techniques and methods proposed to detect falls and anomalies in movements of the elderly, through monitoring of their daily life activities. These methods can be broadly divided into three main categories: 1) Video Analysis Based; 2) Acoustic and Ambience Sensor Based; and 3) Kinematic Sensor Based. This paper critically analyzes the various proposed methodologies, comparing their strengths and weaknesses. We further propose our own technique for fall detection and monitoring of common daily life activities (walking, running, sitting, standing, and lying down), through a novel approach that provides a low cost solution and ensures the safety and security of the elderly without restricting them to confined surroundings.