Iris Segmentation Using Polygon Vertex Curvature
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing iris segmentation methods face challenges with occlusions from eyelashes, eyelids, and lighting distortions, requiring complex algorithms and precise manual annotations, which are time-consuming and not scalable.
Innovation Solution
A method and system for iris segmentation using a trained object detection model to generate polygons on the iris and pupil areas, identifying internally and externally exposed vertices, constructing a closed curvature of the iris region, and segmenting the iris by removing the pupil region.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deep learning based solutions such as U-nets and VGG-Net are used for iris segmentation, then segmentation accuracy is improved, but the requirement for precise manual annotations increases complexity and time consumption
Solution Approach 1:
The system uses self-trained object detection models that automatically detect iris regions without requiring manual annotations. The model trains on unannotated data and automatically generates the necessary segmentation information, eliminating the need for time-consuming manual annotation processes while maintaining high segmentation accuracy.
Solution Approach 2:
The patent creates synthetic annotated data by generating polygons and vertices from detected iris regions. These synthetic annotations are then used to train the object detection model, allowing the system to learn from automatically generated data rather than requiring manual annotations for training.
2Reliability
If complex algorithms are used to remove occlusions from eyelashes and eyelids, then segmentation reliability is improved, but processing time increases
Solution Approach 1:
The system segments the iris region by detecting polygonal boundaries and vertices that define the iris area. By breaking down the complex task of occlusion removal into simpler geometric segmentation steps, the system achieves reliable segmentation without requiring complex time-consuming algorithms to handle eyelashes and eyelids.
Solution Approach 2:
The patent replaces complex algorithmic approaches with geometric and mathematical operations. Instead of using complex image processing algorithms to remove occlusions, the system uses polygon detection, vertex identification, and geometric calculations to accurately segment the iris region, significantly reducing processing time while maintaining reliability.
3Measurement precision
If manual annotation tools are used for iris segmentation, then annotation precision is improved, but scalability is reduced due to human intervention requirements
Solution Approach 1:
The system performs self-service by automatically detecting and annotating iris regions without human intervention. The object detection model processes images autonomously, generating precise annotations and segmenting the iris region, thereby eliminating scalability limitations associated with manual annotation tools while maintaining high precision.
Solution Approach 2:
The patent changes the fundamental parameter of annotation generation from manual human input to automated computational output. By transforming the annotation process from a manual task to an automated algorithmic process, the system achieves both high precision and scalability, as the same algorithm can process unlimited numbers of images consistently.
Data Source
AI summary
In the Iris based biometric recognition systems, iris needs to be segmented before comparison and hence iris segmentation is a crucial process in iris based biometric recognition systems. Existing techniques are unable to segment iris due to occlusions and require complex and time consuming algorithms to remove these occlusions. Other deep learning based solutions require precise annotations of the iris regions along with eyelids and eyelashes. The present disclosure initially generates polygons on the iris and pupil area using a pre-trained detection model. Internally covered vertices and externally exposed vertices of the generated polygons are computed to construct curvature of the iris region. A pupil region is generated by filtering the plurality of vertices lying inside pupil circle region from the plurality of externally exposed vertices. Further, a segmented iris image is generated by removing the pupil region from the input image based on the curvature of the iris region.