Walkthrough Biometric Authentication Using Gait-to-Iris Candidate Narrowing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing walkthrough authentication systems require multiple cameras and hardware resources to perform high-speed face and iris authentication on moving individuals, limiting the time available for authentication and increasing the accuracy of the authentication process, and requiring multiple cameras and hardware resources.
Innovation Solution
A biometric authentication device that performs gait or human shape authentication near the entrance and iris authentication near the exit, using a single camera to capture entire body and eye images, narrowing candidates for iris authentication, thus reducing hardware requirements and enhancing accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple cameras are used to perform both face authentication and iris authentication simultaneously, then authentication accuracy is improved, but device complexity and hardware resources are increased
Solution Approach 1:
The authentication process is divided into two sequential stages: first, face authentication is performed using a camera positioned at the entrance; second, iris authentication is performed using the same camera positioned at the exit. This temporal and spatial segmentation allows a single camera to perform multiple authentication functions without requiring simultaneous multi-camera coverage, thus reducing hardware complexity while maintaining authentication accuracy through multi-modal verification.
2Reliability
If face authentication and iris authentication are performed simultaneously on moving users, then authentication reliability is improved, but the time required for authentication increases
Solution Approach 1:
Face authentication is performed as a preliminary action at the entrance before the user completes their movement through the authentication zone. The system captures facial images and extracts feature amounts in advance, then stores these for later comparison. When the user reaches the exit, the pre-extracted facial features are quickly compared against stored data, and only after successful face authentication does the system proceed to iris authentication. This preliminary processing reduces the time required for simultaneous authentication by preparing data in advance.
3Device complexity
If a single camera is used to capture both entire body and eye images, then hardware resources are reduced, but image quality and authentication accuracy may deteriorate
Solution Approach 1:
The system dynamically adjusts the camera's imaging characteristics based on the authentication stage and user position. When capturing entire body images at the entrance, the camera uses settings optimized for capturing full-body silhouettes and gait patterns. When capturing eye images at the exit, the camera switches to settings optimized for high-resolution眼部 detail capture. This dynamic adaptation of imaging parameters allows a single camera to maintain high image quality across different authentication requirements without requiring multiple specialized cameras.
Data Source
Figure 1
Figure 2~3
Figure 4
AI summary
An authentication device or the like capable of executing biometric authentication with high accuracy on an authentication target who is moving with less hardware resources and within a predetermined time period is provided. A biometric authentication device 100 includes a detection unit 11 that detects an authentication target who is moving in a predetermined direction, a first image input unit 12 that inputs a first image in which an entire body of the target is imaged, a first storage unit 15 that stores characteristic information of the entire body of the target to be authenticated, a first authentication unit 14 that compares the characteristic information calculated from the first image with the characteristic information stored in the first storage unit 15 and extracts candidate information from the first storage unit 15, a second image input unit 13 that inputs a second image obtained by imaging an iris of at least one left and right eyes of the moving target, a second storage unit 17 that stores characteristic information of the irises of the left and right eyes of the target to be authenticated, and a second authentication unit 16 that compares characteristic information of the iris of the second image with the characteristic information stored in the second storage unit 17 for each target included in the candidate information and authenticates the target in the second image.