A Low-Complexity Compressed Sensing Reconstruction Method for Heart Signal Biometric Recognition.

Area of Science:

• Biometrics and Signal Processing
• Wearable Technology and Health Monitoring

Background:

• Biometric systems are crucial for secure authentication, driving research in efficient data acquisition.
• Compressed sensing offers a synchronous data sampling and compression method ideal for resource-constrained portable devices.

Purpose of the Study:

• To develop a compression reconstruction method for bioelectric signals using compressed sensing.
• To improve the efficiency of portable remote bioelectric signal recognition equipment.

Main Methods:

• Utilized electrocardiograms (ECGs) and photoplethysmograms (PPGs) as research data.
• Developed an improved segmented weak orthogonal matching pursuit (OMP) algorithm for signal compression and reconstruction.
• Extracted feature values from reconstructed signals for identification.

Main Results:

• Achieved high recognition rates: 95.65% for ECG and 91.31% for PPG.
• Demonstrated a low residual error of less than 0.05 mV for reconstructed signals.
• Validated the effectiveness of compressed sensing for bioelectric signal compression reconstruction.

Conclusions:

• The proposed compressed sensing method is effective for bioelectric signal compression reconstruction.
• This approach enhances the capabilities of wearable devices for cardiac signal identification.
• The method offers a practical solution for secure and efficient remote biometric recognition.