Cascaded Eye Shape Model for Robust Biometric Feature Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing biometric technologies face challenges in accurately extracting iris features from eye images due to occlusions by eyelids, poor resolution, and varying angles, which affect gaze estimation and identification applications.
Innovation Solution
A detailed eye shape model is calculated using cascaded shape regression methods, estimating the shape of the pupil, iris, and eyelids to enhance feature detection and reduce noise, allowing for more robust biometric applications such as gaze estimation and identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional biometric methods search the entire eye image to detect iris features, then comprehensive feature detection is possible, but processing time and computational complexity increase significantly
Solution Approach 1:
The patent segments the eye image into distinct anatomical regions (pupil, iris, sclera, eyelids) using a pre-established eye shape model. This segmentation allows the system to focus feature detection only within the iris region boundaries, eliminating the need to search the entire eye image while maintaining comprehensive iris feature detection.
Solution Approach 2:
The patent performs preliminary estimation of the eye shape model (including pupil and iris boundaries) before conducting iris feature detection. This preliminary action defines the search region in advance, allowing subsequent feature extraction to be confined to the estimated iris area, thereby reducing processing time without compromising detection accuracy.
2Adaptability or versatility
If eye images are captured at varying angles or with occlusions, then real-world applicability is maintained, but feature extraction accuracy deteriorates
Solution Approach 1:
The patent employs cascaded shape regression that adapts to varying eye image parameters including rotation angles, scale variations, and occlusion conditions. The regression model adjusts shape parameters dynamically based on the input image characteristics, maintaining accurate iris boundary estimation even when images are captured at different angles or partially occluded by eyelids.
Solution Approach 2:
The patent introduces an intermediate eye shape model (comprising pupil and iris boundaries) that mediates between the raw eye image and the final feature extraction process. This intermediate model serves as a robust reference framework that remains accurate even under varying capture conditions, enabling reliable feature extraction despite angle variations or occlusions.
3Measurement precision
If the eye shape model includes detailed boundaries (pupil, iris, eyelid), then feature detection precision improves, but model complexity and training requirements increase
Solution Approach 1:
The patent segments the eye shape estimation into distinct anatomical components (pupil boundary, iris boundary, eyelid contours) that can be modeled and trained independently. This segmentation allows the system to achieve high boundary detection precision for each component while managing overall model complexity through modular training approaches.
Solution Approach 2:
The patent implements a dynamic cascaded regression framework where the eye shape model is iteratively refined through multiple regression stages. Each stage adjusts shape parameters dynamically based on previous estimates, allowing the model to achieve high precision boundaries adaptively without requiring an excessively complex static model structure.
Data Source
AI summary
Systems and methods for robust biometric applications using a detailed eye shape model are described. In one aspect, after receiving an eye image of an eye (e.g., from an eye-tracking camera on an augmented reality display device), an eye shape (e.g., upper or lower eyelids, an iris, or a pupil) of the eye in the eye image is calculated using cascaded shape regression methods. Eye features related to the estimated eye shape can then be determined and used in biometric applications, such as gaze estimation or biometric identification or authentication.