Elcare: Elderly Care With Fall Detection

Abstract

Abstract Continuous growth in the aged population over the years around the globe suggests advancements in medical facilities to ensure good medical health of people. The elder population is vulnerable to many medical problems as the immune system as well as the musculoskeletal system weakens with time. Falls are the major cause leading to serious medical conditions such as paralysis or even death for elders. Early fall detection can help in reducing the severity of these accidents and provide immediate medical assistance. To detect falls this paper discusses the use of pose estimation and sensor-based mobile devices. The system makes use of MediaPipe’s Pose Detection method to form a skeletal structure of the person. Exploiting the features such as drastic change in coordinate axes, angle of inclination while falling, and the sudden change in velocity during a fall helps to verify the authenticity of the fall. The sensor-based method makes use of mobile sensors and records the change in their values to decide whether the activity is a fall or not. The combination of both these methods allows for an accurate fall detection mechanism where the system finally notifies the concerned authorities via real-time feedback techniques such as text, message, or push notification.