Fall Detection for the Elderly Based on 3-Axis Accelerometer and Depth Sensor Fusion with Random Forest Classifier.

Area of Science:

• Biomedical Engineering
• Computer Science
• Robotics

Background:

• Falls are a significant health risk, especially for the elderly.
• Existing fall detection systems often suffer from high false detection rates.
• Integrating multiple sensor modalities can improve accuracy.

Purpose of the Study:

• To develop a robust and accurate fall detection method.
• To minimize false positives in fall detection.
• To create a more realistic dataset for fall detection research.

Main Methods:

• A novel fall detection system combining 3-axis accelerometer and depth sensors.
• Feature extraction from both vision and acceleration data.
• Machine learning algorithms for classification and generalization.
• Development of a new, realistic fall event database.

Main Results:

• The proposed method significantly reduces the false detection rate compared to single-feature methods.
• The combined sensor approach achieves good generalization performance.
• Experimental results demonstrate efficient and accurate fall detection.

Conclusions:

• Combining accelerometer and depth sensor data enhances fall detection accuracy.
• Machine learning integration provides robust generalization.
• The new database facilitates more realistic fall detection research.