A Novel Fall Detection System using Mediapipe

The society values the elderly, yet because of their deteriorating health as they age, they are more susceptible to serious injury from falls. This study describes a mediapipe-based vision-based fall detection system that notifies caretakers by chatbot when a fall is detected. The camera is placed so that the area in which the elderly spend most of their time receives the most coverage. The Mediapipe architecture is taught to identify a fall, and the Random Forest approach is used to categorize diverse circumstances. The model has a 90% accuracy rate and is designed to identify a single person at a time.