Fall Detection and Notification System for the Elderly using a fusion of Geophone and Thermal Imaging Camera sensors.

Falls among elderly individuals living alone pose significant risks to their well-being. However, conventional fall detection devices often compromise privacy or are easily forgotten or rejected by the users. To address this issue, we present a combined fall detection device utilizing a Geophone as the primary sensor and a Thermal Imaging camera as the secondary sensor. The device aims to provide timely alerts to caregivers while respecting the privacy concerns of the elderly. Our approach utilizes the ESP32 microcontroller, Raspberry Pi, and incorporates a Thermal Sensor (MLX90640) and Geophone (SM-24 10Hz). This paper describes the design, implementation, and evaluation of our non-wearable fall detection device, highlighting its potential to enhance the safety and privacy of elderly individuals living alone. Our novelty is implementing a non-wearable device with sensor fusion. Wearable devices can cause discomfort and inconvenience for elderly individuals. Moreover, there are instances where elderly individuals may wake up during the night and forget to put on the wearable device, leaving them vulnerable to fall incidents without proper monitoring or assistance.