Radar Signal Data for Detecting Living Bodies in Vehicles

A novel radar signal processing technique uses mm-wave frequency-modulated continuous-wave radar to identify living bodies in vehicles by monitoring the Doppler effect from consistent breathing. The method was assessed with two phantoms mimicking child breathing and with adult subjects in various conditions. The algorithm, developed by Hajar Abedifirouzjaei at the University of Waterloo, is described as robust, accurate, and fast, requiring no training or threshold definition.

Use Cases

• Developing in-vehicle occupant detection systems based on the described Doppler monitoring of breathing signals.
• Benchmarking new signal processing algorithms against the described constant false alarm rate (CFAR)-free method.
• Validating radar-based safety systems for detecting children or pets left in cars using the described phantom and adult test data.

Strengths

• The method was validated with two distinct phantoms mimicking child breathing and with adult subjects.
• The algorithm is described as not requiring training or threshold definition, which suggests ease of deployment.
• The system's performance was assessed over multiple months, suggesting long-term testing.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count and dataset size are unknown, which may limit suitability assessment.
• The raw data and specific performance metrics are not provided, only the method description.

Provenance

Source

University of Waterloo

Collection Method

Experimental assessment using radar signals from phantoms and human subjects in vehicles.