Multi-Stage Iris Segmentation for Handheld Device Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Iris recognition systems face challenges in efficiently detecting and segmenting irises with varying sizes and locations, especially in handheld device images where distance and image quality can lead to slow processing and false candidates.
Innovation Solution
A method combining eye detection, eye modeling, and iris segmentation, which includes multiple stages of iris identification and segmentation with confidence level checks to ensure accurate iris code extraction, even in images with high variation in iris size and location, using techniques like Haar classifiers, active appearance models, and neural networks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If large-scale eye detection is performed before iris detection to speed up processing, then processing speed is improved, but segmentation accuracy deteriorates in cases where only a portion of the eye is included in the image
Solution Approach 1:
The detection process is segmented into multiple stages: first performing large-scale eye detection to locate the general eye region, then performing refined iris detection specifically within that detected eye region. This multi-stage segmentation approach allows the system to benefit from the speed of large-scale detection while ensuring accurate iris segmentation through the subsequent refined detection stage.
Solution Approach 2:
The system performs preliminary eye detection to identify the eye region before conducting the main iris detection and segmentation. This preliminary action narrows down the search space and provides a foundation for more accurate subsequent processing, while confidence level checks ensure that only reliable detections proceed to the next stage.
2Measurement precision
If iris detection is performed on the whole acquired image to ensure accurate location, then detection accuracy is improved, but processing time increases
Solution Approach 1:
The image processing is segmented into stages: first performing rapid large-scale eye detection on the whole image to locate the eye region, then performing detailed iris detection only within that specific eye region. This segmentation reduces the computational burden compared to searching the entire image, while maintaining accuracy through the focused refined detection.
Solution Approach 2:
The system extracts the eye region from the whole acquired image through preliminary eye detection, then performs iris detection only on this extracted region. This extraction approach eliminates the need to process the entire image, significantly reducing detection time while maintaining accuracy through the subsequent refined iris detection on the extracted eye region.
3Reliability
If multiple stages of iris identification and segmentation are performed with confidence level checks, then recognition reliability is improved, but system complexity increases
Solution Approach 1:
The system performs preliminary iris identification and segmentation with confidence level checks before final recognition. This preliminary action filters out low-quality detections early in the process, ensuring that only high-confidence results proceed to final recognition, thereby improving reliability without requiring complex post-processing.
Solution Approach 2:
The system implements feedback through confidence level checks at multiple stages. The confidence levels provide feedback on the quality of detected irises, allowing the system to accept or reject detections based on predefined thresholds. This feedback mechanism improves recognition reliability by ensuring only high-quality detections are used, while the automated threshold-based decisions keep the system complexity manageable.
Data Source
AI summary
A method of iris recognition comprises detecting a body region larger than and comprising at least one iris in an image and performing a first eye modelling on the detected body region. If successful, the result of first iris segmentation based on the first eye model is chosen. Otherwise, a first iris identification is performed on the detected body region. If successful, the result of second iris segmentation based on a second eye modelling is chosen. Otherwise, second iris identification is performed on the image, third eye modelling is performed on the result of the second iris identification, and third iris segmentation is performed on the result of the third eye modelling. If successful, the result of third iris segmentation based on a third eye modelling is chosen. An iris code is extracted from any selected iris segment of the image.