Spoof Detection via Specular and Diffuse Reflection Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Optical fingerprint sensors are vulnerable to spoofing, as they can be easily deceived by two-dimensional representations, and there is a need for improved spoof detection methodologies to enhance their reliability.
Innovation Solution
The method employs a color controllable light source with varying light intensity patterns to capture images under both specular and diffuse illumination, using a photodetector pixel array to extract metrics that differentiate between live fingers and spoofs based on their optical properties, including specularity and color information, and utilizes machine learning algorithms for authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If optical fingerprint sensors are used for biometric authentication, then convenience and security are improved, but vulnerability to spoofing increases
Solution Approach 1:
The patent segments the illumination into multiple distinct light patterns with different properties (specular, diffuse, structured lighting). By capturing images under these different illumination conditions separately and analyzing them independently, the system can identify spoofing attempts that would be invisible under uniform illumination, thereby resolving the contradiction between authentication reliability and spoofing vulnerability
Solution Approach 2:
The patent changes illumination parameters (light intensity distribution, polarization state, wavelength) to create multiple imaging conditions. By varying these parameters and comparing the resulting images, the system can detect subtle differences between real fingerprints and spoofs, improving reliability while maintaining resistance to spoofing attacks
2Measurement precision
If multiple illumination patterns are used to improve spoof detection, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent employs a single light source that can generate multiple illumination patterns (specular, diffuse, structured) by modulating its properties. This multi-functional approach allows the system to achieve high spoof detection precision without requiring multiple separate illumination devices, thereby resolving the contradiction between measurement precision and device complexity
Solution Approach 2:
The patent uses periodic modulation of the light source to sequentially apply different illumination patterns. By timing the capture of images under each illumination condition appropriately, the system achieves high detection precision while using a single controllable light source, reducing overall system complexity
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This approach effectively distinguishes between live fingers and spoofs by analyzing image portions under different illumination conditions, significantly improving the reliability of biometric authentication by leveraging the unique optical behaviors of live fingers and spoofs.
Implementation Method 1
the image is dominated by specular reflections from the object, originating from the light source unit above the pixels capturing the image portion
Implementation Method 2
the image is dominated by diffuse reflections from the object originating from light source units adjacent to the at least partly in-active illumination area
Data Source
AI summary
The present invention relates biometric authentication using an optical biometric arrangement comprising an image sensor comprising a photodetector pixel array configured to capture an image of an object, the image sensor being arranged under a color controllable light source comprising light source units, the method comprising: providing a light pattern comprising portions of different light intensity for illuminating the object; acquiring an image of the object, the image comprising image portions corresponding to the portions of different light intensity of the light pattern illuminating the object, at least one image portion being captured by pixels in the photodetector pixel array arranged directly under a light source being active during image acquisition, and at least one image portion being captured by pixels in the photodetector pixel array arranged under an at least partly in-active illumination area of the color controllable light source during image acquisition, and performing biometric authentication at least partly based on metrics extracted from the image portions.