Living-Skin Detection Based on Spatio-Temporal Analysis of Structured Light Pattern.

Area of Science:

• Biomedical Optics
• Computer Vision
• Biometrics

Background:

• Living-skin detection is crucial for applications like imaging photoplethysmography and biometric anti-spoofing.
• Existing methods may not fully leverage the complex optical properties of skin.
• Structured light offers a promising avenue for enhanced detection capabilities.

Purpose of the Study:

• To develop a novel living-skin detection algorithm utilizing spatio-temporal characteristics of structured light.
• To differentiate between living skin and non-skin surfaces based on laser spot interactions.
• To improve the accuracy and robustness of biometric and physiological imaging systems.

Main Methods:

• Exploited spatio-temporal features of structured laser patterns projected onto surfaces.
• Analyzed frequency-domain sharpness differences due to laser-tissue interactions in multi-layer skin.
• Incorporated brightness fluctuations caused by physiological motion and laser interference on living skin.

Main Results:

• The proposed method achieved high performance with an average precision of 85.32%, recall of 83.87%, and F1-score of 83.03% across different environments.
• The hybrid approach, combining frequency-domain sharpness and brightness fluctuations, showed significant improvements over methods using only skin structure properties.
• Demonstrated efficacy in both controlled dark chamber experiments and a clinical Neonatal Intensive Care Unit (NICU) setting.

Conclusions:

• The integration of frequency-domain sharpness and brightness fluctuation analysis effectively enhances living-skin detection.
• Structured light-based methods offer a robust and accurate approach for distinguishing living skin from other materials.
• This research provides a strong foundation for advanced structured light-based physiological imaging and anti-spoofing technologies.