Project Overview: This project presents a real-time fall detection system that utilizes LiDAR sensors to accurately classify human activities and identify potential falls among elderly people. The system achieved an accuracy of 84% in Human Activity Classification, demonstrating the potential for supporting elderly care with prompt fall detection.
- Real-time fall detection with an accuracy of 84%.
- Extensive dataset created using LiDAR sensors, featuring over 40+ diverse scenarios and simulated falls.
- Coordinated with a team of 9 to develop a supervised Deep Learning model specifically for fall detection using LiDAR sensor data.
This system leverages LiDAR sensors to capture and process real-time data, enabling accurate activity classification and fall detection. The robust dataset, developed through extensive simulation, enhances the model's reliability across a wide range of fall scenarios.
This work was published in the IEEE Journal as part of the paper titled "Fall Detection for Elderly People using LiDAR Sensor" and was presented at the 3rd IEEE International Conference on Artificial Intelligence for Internet of Things (AI-IoT) held at VIT, Vellore in 2024.
- Viswa - av.ambaram06@gmail.com
- Dharma Pravardhana - dharmapravardhana7@gmail.com