Chrominance-Based Face Liveness Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional facial recognition systems are vulnerable to face spoofing techniques, such as using printed photographs or videos, due to inadequate liveness detection mechanisms, which can be easily bypassed, compromising security and user convenience.
Innovation Solution
A method and system utilizing multi-color illumination to differentiate between 2D and 3D facial images by capturing images under different color conditions and analyzing chrominance changes to determine if the facial image is consistent with a 3D structure, thereby authenticating the user without requiring user interaction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional liveness detection methods are used, then the system is simple to implement, but the security against face spoofing is insufficient
Solution Approach 1:
The patent uses multi-color illumination to capture facial images under different color conditions. By analyzing chrominance changes across multiple color illuminations (e.g., red, green, blue lights), the system can distinguish between real 3D faces and 2D spoofing attempts. Different colors interact differently with skin tissue, creating characteristic chrominance patterns that reveal liveness.
Solution Approach 2:
The patent transitions from traditional single-color or grayscale imaging to multi-color chrominance analysis. By adding the chrominance dimension (analyzing color information across multiple wavelengths), the system creates additional discrimination capability that helps differentiate between 2D and 3D structures without requiring complex hardware modifications.
2Reliability
If challenge-response methodology is used for liveness detection, then user interaction is required, but this causes inconvenience to end users
Solution Approach 1:
The system performs automatic liveness detection by analyzing chrominance patterns in facial images captured under multi-color illumination. No user action or response is required - the detection happens passively as the user simply presents their face for authentication. The system self-determines liveness based on the optical properties of real skin versus 2D materials.
3Reliability
If traditional username/password authentication is used, then it is easy to implement, but it loses ability to provide secure authentication
Solution Approach 1:
The patent enhances traditional face recognition by incorporating multi-color chrominance analysis. The system captures facial images under different color illuminations and analyzes the chrominance characteristics, which are unique to real human skin. This adds a layer of optical verification to the authentication process, making it resistant to photo and video spoofing while building upon existing face recognition infrastructure.
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 provides a robust and reliable liveness detection mechanism that effectively counters various face spoofing techniques, ensuring high assurance and seamless user experience while maintaining security.
Implementation Method 1
acquiring a first image of the facial image illuminated with a first color, acquiring a second image of the facial image illuminated with a second color
Implementation Method 2
a camera acquires images in at least the first direction
Data Source
AI summary
Methods, systems and computer readable medium for liveness detection authentication of a facial image are provided. The method includes acquiring a first image of the facial image illuminated with a first color, acquiring a second image of the facial image illuminated with a second color, and determining if the facial image is consistent with a three-dimensional (3D) structure in response to a combination of the first and second images. The method further includes authenticating the facial image if the facial image is consistent with a 3D structure and the facial image matches a face of user to be authenticated.