Real-Time Monitoring and Human Fall Detection using Hybrid Neural Networks (HNN)
S R Varshini, Ravina Padmaja B S, J. Gnanasoundharam · 2025
This research work develops a system that detects falls in real time for old people using Hybrid Neural Networks (HNNs). Falls mainly to the elderly are crucial because they expose the patients or even lead to death as it causes serious health impacts. The system uses one camera to monitor posture or movements, which then analyzes pattern motions to correctly detect falls accurately. The system is trained on a dataset of fall and non-fall activities, which enables it to improve its detection accuracy continuously. Besides monitoring falls, the system also caters to people with mental disabilities and those who are prone to self-harm, thus ensuring that critical situations are addressed promptly. The approach is a combination of advanced computer vision and deep learning techniques that enable continuous, automated fall detection to improve the safety and well-being of vulnerable individuals. As a result, the system offers a reliable mechanism for immediate assistance, thus enhancing healthcare services and supporting individuals in maintaining independence with reduced risks of falls and emergencies.