An Novel Based Mobile Dropping Sesing to Protect using Inertia Sensor and Integration of DL Technology

With all such technologies, the recent advances in health monitoring due to smartphone sensor technologies provide an opportunity to quantify an individual’s functional performance and deviations from normality. Falling and drowning are two major unnatural reasons for accidental silent death, and it requires an ambient approach to detect the same. This paper presents a novel ambient assistive framework, FaDD (Falling and Drowning Detection), designed for identifying and distinguishing incidents of falls or drowning. Input data are taken through smartphone sensors like the accelerometer, gyroscope, magnetometer, and GPS, which capture the movement of a person’s body for effective recognition. By adopting a hierarchical approach, FaDD effectively detects dropping and choking actions using machine learning models. The framework is also integrated with the emergency alert notifications embedded in a smartphone application and, in the case of drowning, immediately alerts the respective stakeholders of the guardians, rescue team, and close community circle. With an accuracy rate of ${9 8 \%}$, FaDD is proficient in detecting falls, drowning, and normal activities. Coordination of all stakeholders through the FaDD framework is bound to enhance the provision of an effective and efficient health service to the people.