RaspiReader: Open Source Fingerprint Reader

Area of Science:

• Biometrics
• Computer Vision
• Hardware Security

Background:

• Developing cost-effective and accessible biometric hardware is crucial for research.
• Existing optical fingerprint readers often lack advanced spoof detection capabilities.
• Prototyping new fingerprint sensing technologies can be time-consuming and expensive.

Purpose of the Study:

• To introduce RaspiReader, an open-source, low-cost, and easily assembled optical fingerprint reader.
• To facilitate rapid prototyping of novel fingerprint sensing and spoof detection techniques.
• To demonstrate an improved hardware-based spoof detection method using multi-modal imaging.

Main Methods:

• Designed and open-sourced RaspiReader using ubiquitous components, achievable in under an hour for $175.
• Customized RaspiReader with two cameras: one for frustrated total internal reflection (FTIR) imaging, another for direct finger imaging.
• Developed a spoof detection technique by fusing complementary information from both image streams.

Main Results:

• RaspiReader provides high-resolution, spoof-resistant fingerprint images.
• The fused image data significantly improved spoof detection performance compared to single-source FTIR.
• Fingerprint matching experiments confirmed interoperability with commercial off-the-shelf (COTS) optical readers.

Conclusions:

• RaspiReader offers a valuable, accessible platform for fingerprint research and development.
• The dual-camera approach provides a robust method for enhancing fingerprint spoof detection.
• Open-sourcing hardware accelerates innovation in biometrics and security research.