Abnormal human activities recognition : brief synthesis of vision based fall detection

Abstract

Various approaches of human activities recognition by vision in the context of elderly care have been proposed in the literature, particularly abnormal activities detection and analysis. This is due to the necessity to note that the observation of changes in the daily activities of individuals, enhanced and instrumental daily life activities are essential to detect medical conditions before they become critical. Taking care of the elderly has an obvious effect on their independence, their psychological status and their quality of life. Unfortunately, those proposals still ineffective and limited, due to the difference in lifestyle, habits and manners of performing a certain activity as well as the variety of activities from one field of application to another. Therefore, we present in this paper a brief synthesis and analysis of some existing works of fall detection.