Extraction of Human Limbs Based on Micro-Doppler-Range Trajectories Using Wideband Interferometric Radar.

Area of Science:

• Radar Signal Processing
• Human Motion Analysis
• Biomedical Engineering

Background:

• Accurate human limb motion extraction is crucial for advanced radar-based human motion detection.
• Overlapping limb movements in the time-Doppler plane pose significant challenges for traditional radar analysis.
• Identifying specific body parts contributing to radar signals remains difficult.

Purpose of the Study:

• To develop a method for separating and classifying motions of different human limbs using radar.
• To overcome the limitations of overlapping signals in radar-based human motion analysis.
• To enhance the precision of human motion detection systems.

Main Methods:

• Utilizing interferometric radar to generate micro-Doppler-Range signatures (mDRS).
• Extracting micro-Doppler-Range trajectories (MDRTs) to analyze limb movements.
• Implementing a three-dimensional constant false alarm (3D-CFAR) detection algorithm to resolve signal overlaps.
• Employing a 77 GHz radar with 4 GHz bandwidth for data acquisition.

Main Results:

• Successfully extracted and classified motions from different human limbs.
• Demonstrated the ability to resolve overlapping signal components in the time-Doppler plane.
• Experimental validation using a Kinect sensor confirmed the approach's effectiveness.
• Achieved repeatable and robust results across multiple subjects and trials.

Conclusions:

• The proposed mDRS-based radar technique effectively extracts and classifies individual human limb motions.
• This method significantly improves upon existing radar capabilities for human motion detection by resolving signal ambiguities.
• The approach shows promise for robust and repeatable human motion analysis in various scenarios.