Related Experiment Video
Updated: Aug 13, 2025
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
10.8K
A Non-Contact Fall Detection Method for Bathroom Application Based on MEMS Infrared Sensors
Chunhua He1, Shuibin Liu1, Guangxiong Zhong1
1School of Computer, Guangdong University of Technology, Guangzhou 510006, China.
Micromachines
|January 21, 2023
Summary
A new non-contact fall detector using infrared sensors and AI accurately identifies elderly falls in bathrooms. This low-cost, privacy-preserving Internet of Things (IoT) device offers reliable fall detection for enhanced safety.
Area of Science:
- Gerontology and Health Technology
- Artificial Intelligence in Healthcare
- Sensor Technology and Signal Processing
Background:
- The global elderly population is increasing, with falls posing a significant health risk, particularly in bathrooms.
- Existing fall detection methods often lack non-contact capabilities, privacy, or affordability.
- There is a critical need for effective, unobtrusive fall detection systems for the elderly.
Purpose of the Study:
- To design and evaluate a non-contact fall detection system specifically for elderly individuals in bathroom environments.
- To develop and implement advanced image processing algorithms for accurate fall event identification.
- To assess the feasibility of using Micro-electromechanical Systems Pyroelectric Infrared (MEMS PIR) and thermopile IR array sensors for fall detection.
Main Methods:
- Development of a non-contact fall detector integrating MEMS PIR and thermopile IR array sensors.
- Implementation of image processing techniques including low-pass filtering and double boundary scans.
- Extraction of statistical features (area, center, duration, temperature) and classification using a 3-layer BP neural network.
Main Results:
- The system achieved high performance metrics: precision (94.45%), recall (90.94%), detection accuracy (92.81%), and F1-Score (92.66%).
- The detection method demonstrated feasibility across various conditions, including different ambient temperatures, lighting, and fall scenarios.
- The developed Internet of Things (IoT) detector is validated as effective, low-cost, and privacy-secure.
Conclusions:
- The novel non-contact fall detection system using MEMS PIR and thermopile IR sensors is highly effective for bathroom environments.
- The integration of image processing and BP neural networks provides robust fall event identification.
- This affordable and privacy-guaranteed IoT solution holds significant potential for widespread household use in elderly care.
