Sweet-An Open Source Modular Platform for Contactless Hand Vascular Biometric Experiments

Area of Science:

• Biometrics and Pattern Recognition
• Computer Vision
• Human-Computer Interaction

Background:

• Traditional vein recognition systems require physical contact, posing hygiene challenges.
• Contactless biometrics is crucial for applications demanding high sanitation standards.
• Existing contactless solutions for vascular biometrics are limited.

Purpose of the Study:

• To develop and evaluate a novel contactless vascular biometrics sensor platform.
• To enable multi-modal data acquisition for hand vascular and surface features.
• To create a new dataset for finger-vein recognition research.

Main Methods:

• Developed the 'sweet' contactless sensor platform supporting multi-spectral Near-Infrared (NIR), RGB-color, Stereo Vision (SV), and Photometric Stereo (PS) modalities.
• Collected a comprehensive dataset of finger, palm, and wrist vascular data from 120 subjects.
• Conducted experiments focusing on Finger-Vein Recognition (FVR) and explored multi-modal fusion techniques.

Main Results:

• The contactless platform successfully acquired vascular and surface data from fingers, palms, and wrists.
• Experimental results demonstrated the potential of the platform for biometric applications, particularly FVR.
• Analysis included the effectiveness of fusing data from multiple acquisition modalities.

Conclusions:

• The 'sweet' platform offers a viable contactless solution for vascular biometrics, addressing hygiene concerns.
• The newly collected dataset and open-source resources will advance research in contactless biometrics.
• Multi-modal fusion shows promise for enhancing the accuracy and robustness of biometric systems.