Eye Spoof Detection Using Dual Illumination Specular Reflection Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Person detection, recognition, and monitoring systems are vulnerable to spoofing, where images of individuals can be presented instead of real persons, leading to false alarms and failure to detect unauthorized individuals.
Innovation Solution
The method involves illuminating an eye with radiation from two spaced-apart sources, acquiring images of specular reflections, and determining the difference between these reflections to differentiate between real and spoofed eyes by analyzing the distance between reflection positions, generating a data signal based on whether the difference is below or above a threshold.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional imaging systems are used for person detection, then the system can detect and recognize individuals, but the system becomes vulnerable to spoofing attacks where photographs or fake images can circumvent detection
Solution Approach 1:
The illumination system is segmented into multiple independent light sources positioned at different locations. Each illumination source creates distinct specular reflection patterns that can be individually analyzed. This segmentation allows the system to differentiate between real eyes (which produce consistent reflection patterns across multiple illumination angles) and spoofed images (which produce inconsistent or absent reflections), thereby improving spoof detection accuracy without requiring a single complex illumination mechanism.
Solution Approach 2:
The system changes the parameter of illumination angle by using multiple light sources positioned at different locations relative to the eye. By varying the illumination angle parameter, the system captures specular reflections from different perspectives. Real eyes exhibit characteristic reflection patterns that change predictably with illumination angle, while photographs or fake images do not. This parameter variation enables reliable differentiation between authentic and spoofed subjects.
2Measurement precision
If multiple illumination sources are used to detect specular reflections, then spoof detection accuracy improves, but the number of image regions and data processing requirements increase
Solution Approach 1:
The system extracts only the specific specular reflection regions from the captured images, isolating the relevant information needed for spoof detection. By focusing computational resources on analyzing only these extracted reflection regions rather than processing entire images, the system achieves precise eye curvature measurement while minimizing data processing complexity. The extraction process identifies and isolates the characteristic reflection patterns that indicate real versus fake eyes.
Solution Approach 2:
The system performs preliminary identification and segmentation of specular reflection regions before conducting detailed analysis. By pre-processing the images to locate and mark the reflection regions of interest, the system prepares the data in advance for more efficient processing. This preliminary action reduces the computational burden of subsequent analysis by focusing only on the critical reflection areas rather than analyzing entire images.
3Reliability
If the system analyzes specular reflections from multiple illumination positions, then the ability to differentiate real eyes from spoofed images improves, but the time required for detection increases
Solution Approach 1:
The illumination sources are activated in a periodic or sequential manner rather than simultaneously, with each source illuminating the eye in turn. This periodic activation allows the system to capture multiple reflection patterns over time and compare them. Real eyes produce consistent reflection characteristics across periodic illumination cycles, while spoofed images show inconsistencies or lack expected reflection patterns. This periodic approach maintains detection reliability while enabling time-based verification.
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
Effectively detects real eyes and prevents spoofing by accurately determining the curvature of the eye surface, ensuring reliable person detection and monitoring systems.
Implementation Method 1
a first set of image information corresponding to a first image region representing a first specular reflection at a third position relative to the eye, and (b) a second set of image information corresponding to a second image region representing a second specular reflection at a fourth position relative to the eye
Data Source
AI summary
The invention enables spoof detection in eye based person detection, person recognition, or person monitoring systems. The invention includes (i) illuminating an eye from a first source located at a first position, (ii) illuminating the eye from a second source located at a second position spaced from the first position, (iii) acquiring at the image sensor, a set of images of the eye which includes (a) a first set of image information representing a first specular reflection at a third position relative to the eye, (b) a second set of image information representing a second specular reflection at a fourth position relative to the eye, (iv) determining a difference value representing a difference between the third position and the fourth position, and (v) generating a data signal representing detection of a real eye in response to determining that the difference value is less than a threshold value.