Single-Channel FMCW-Radar-Based Multi-Passenger Occupancy Detection Inside Vehicle.

Area of Science:

• Automotive Sensing
• Biomedical Engineering
• Signal Processing

Background:

• Accurate vehicle occupancy detection is crucial for safety and efficiency.
• Existing methods often rely on cameras or complex sensor arrays.
• Radar-based sensing offers a privacy-preserving and robust alternative.

Purpose of the Study:

• To develop and evaluate a novel method for multi-passenger occupancy detection using a single-channel FMCW radar.
• To identify and quantify physiological signals indicative of human presence within a vehicle.
• To assess the efficacy of proposed spectral features for occupancy classification.

Main Methods:

• Utilized a single-channel frequency-modulated continuous-wave (FMCW) radar system.
• Analyzed radar signal time-frequency spectrum to extract physiological characteristics.
• Proposed spectral power and Wiener entropy as key features for detecting breathing and heartbeat.
• Applied clutter removal and feature extraction for seats 1 and 2.
• Employed linear discriminant analysis (LDA) and maximum likelihood estimation (MLE) for classification.

Main Results:

• Spectral power and Wiener entropy effectively quantify physiological movements.
• Feature distributions based on average means of power and Wiener entropy were established.
• Successful classification of multi-passenger occupancy was achieved.
• The proposed features demonstrated high effectiveness for occupancy detection.

Conclusions:

• Single-channel FMCW radar is a viable technology for in-cabin occupancy detection.
• Spectral power and Wiener entropy are promising features for identifying occupants via physiological signals.
• The proposed method offers a robust and potentially cost-effective solution for vehicle occupancy sensing.