Beam Profile Skin Analysis and Depth Checks for 3D Mask Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current face recognition systems are vulnerable to 3D mask attacks and lack ethnic skin agnosticism, leading to security breaches and computational inefficiencies.
Innovation Solution
A method involving face detection, skin detection, and 3D detection using a camera, illumination pattern, and processing unit to authenticate a human face by analyzing geometrical and material properties, distinguishing between real faces and 3D masks, and ensuring agnostic performance across different skin types.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If 3D cameras and multiple video frames are used for presentation attack detection, then detection reliability is improved, but computational power requirements and processing time increase
Solution Approach 1:
The patent extracts and utilizes the natural temporal variations that occur during normal face authentication (head movements, blinking, breathing) to create liveness signals. By focusing on these inherent motions rather than requiring complex multi-frame analysis, the system achieves reliable mask detection with reduced computational overhead.
Solution Approach 2:
The patent introduces an intermediary temporal analysis layer that processes sequential 2D images to detect unnatural rigidity patterns characteristic of masks. This intermediary processing step enables reliable detection without requiring the full computational power of advanced 3D camera systems.
2Measurement precision
If advanced 3D algorithms and multiple video frames are processed, then mask detection accuracy is improved, but power consumption increases
Solution Approach 1:
The patent extracts temporal liveness signals from standard 2D image sequences captured during normal authentication. By utilizing these inherent temporal variations rather than processing intensive 3D data, the system maintains high mask detection accuracy while significantly reducing power consumption.
Solution Approach 2:
The patent uses standard 2D camera images that are already captured for authentication purposes, rather than requiring additional power-intensive 3D sensing. These existing images are repurposed for liveness detection, eliminating the need for expensive additional hardware and reducing overall power consumption.
3Reliability
If skin optical property analysis is performed to detect masks, then detection reliability is improved, but computational demands increase
Solution Approach 1:
The patent segments the authentication process into distinct phases: initial face detection using standard algorithms, followed by temporal liveness analysis of captured images. This segmentation allows skin optical property analysis to be performed efficiently on already-captured images rather than in real-time, maintaining both reliability and speed.
Solution Approach 2:
The patent performs preliminary capture of multiple images during the normal authentication process before mask detection is even considered. This preliminary action ensures that the temporal data needed for reliable mask detection is already available, eliminating the need for additional processing time and maintaining authentication speed.
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
Enhances security by reliably detecting 3D mask attacks and provides efficient, fast authentication across various skin types, reducing computational demands and power consumption.
Implementation Method 1
determining at least one second image using the at least one camera, wherein the second image comprises a plurality of reflection features generated by the scene in response to illumination by the illumination features
Data Source
Figure 1~2
Figure 3
AI summary
A mobile device for face authentication is proposed. The mobile device (124) comprises: a) an illumination unit (118) configure to project at least one illumination pattern comprising a plurality of illumination features on a scene, b) a camera (112) configured to capture a first and a second image of the scene, wherein the second image is captured of the scene under patterned illumination, and c) a processing unit (114) configured to execute: i. a face detection step (110) comprising detecting a face in the first image by identifying at least one pre-defined or pre-determined geometrical feature characteristic for faces; ii. a skin detection step (116) comprising determining a first beam profile information of at least one of the reflection features located inside an image region of the second image corresponding to an image region of the first image comprising the identified geometrical feature by analysis of its beam profile and determining at least one material property of the reflection feature from the first beam profile information, wherein the detected face is characterized as skin if the material property corresponds to at least one property characteristic for skin; iii. a 3D detection step (120) comprising determining a depth map of at least parts of the scene the reflection features located inside the image region of the second image corresponding to the image region of the first image comprising the identified geometrical feature, wherein the detected face is characterized as 3D object if the depth level deviates from a pre-determined or pre-defined depth level of plane objects; a) an authentication step (122) comprising authenticating the detected face if in step ii) (116) the detected face is characterized as skin and in step iii) (120) the detected face is characterized as 3D object.