Adaptive Biometric Recognition Using Iris and Facial Feature Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current biometric authentication technologies face challenges in efficiently and accurately recognizing users, especially when iris recognition quality is invalid, leading to reduced accuracy or failure in authentication processes.
Innovation Solution
A method and apparatus that extract facial feature data from frame images when iris recognition quality is invalid, and perform iris recognition when quality is valid, using a combined score based on iris and facial feature data to determine user recognition, with the ability to accumulate and reuse facial feature data across frames for improved accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iris recognition is performed based on quality information, then authentication accuracy is improved, but when iris quality is invalid, recognition fails or accuracy decreases
Solution Approach 1:
The system dynamically switches between iris recognition and facial feature recognition based on the quality of the iris area. When iris quality is valid, iris recognition is performed; when iris quality is invalid, the system transitions to facial feature recognition, making the recognition method adaptive to changing conditions and maintaining reliability across different quality scenarios
Solution Approach 2:
The system changes the recognition parameter from iris-based to facial-feature-based according to the quality information of the iris area. This parameter switching allows the system to maintain authentication accuracy by selecting the appropriate recognition method based on the current quality conditions
2Productivity
If only iris recognition is used, then the authentication process is simple and fast, but accuracy decreases when iris quality is poor
Solution Approach 1:
The system maintains simplicity and speed by using iris recognition as the default method when quality is valid, but dynamically introduces facial feature recognition only when needed (when iris quality is invalid), thus preserving authentication speed while improving accuracy under poor conditions
Solution Approach 2:
The system extracts and uses only the necessary recognition method based on quality conditions - primarily iris recognition for speed, with facial feature extraction activated only when iris quality is insufficient, avoiding unnecessary processing while maintaining accuracy
3Measurement precision
If facial feature data is extracted and accumulated across frames, then recognition accuracy is improved when iris quality is invalid, but processing complexity increases
Solution Approach 1:
The system performs preliminary extraction and accumulation of facial feature data from multiple frames in advance, storing this data for later use when iris recognition is unavailable. This preliminary preparation ensures that when iris quality is invalid, the accumulated facial feature data can be immediately used without delay, improving accuracy while managing complexity through advance preparation
Solution Approach 2:
The system dynamically adjusts processing complexity by accumulating facial feature data only when necessary (when iris quality is invalid), rather than continuously processing all frames. This conditional accumulation reduces overall processing complexity while maintaining accuracy when needed
Data Source
AI summary
Disclosed is a user recognition apparatus and method, the apparatus configured to determine a validity of an iris area for each frame, to perform an iris recognition when the iris area is valid, and to extract a facial feature for a fusion recognition when the iris area is invalid.