An Intelligent Mobile App for Fall Detection
Mohammad Salama, Khaled Eskaf · 2020
Falling is one of the most prominent external causes of unintentional injury. On the other hand, the number of elderly people living alone has been continuously growing worldwide. This independence comes with the risk of not receiving a prompt attention if an accident occurs. Lack of attention in the first hour of the accident increases the risk of death and chronic affections. In this paper, a real-time, lowcomplexity mobile fall detection and alert system was developed for helping -but not limited to- elderly people. The system uses smartphones as a platform for fall detection. The fall detection approach is based on processing accelerometer data which create a recognizable pattern during a fall. The proposed system uses Neural Network as a classifier for fall detection, and voice recognition for capturing and confirming the need of sending help alerts. Deep Neural Network was also implemented to see its effect on performance. Experimental results show that the system can detect falls with high accuracy (up to 96%) while consuming reasonable amount of power.