Eyeball Image Segmentation for Outdoor Iris Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Iris authentication systems fail under external light conditions due to exposure and reflection, causing inaccuracies in iris code matching.
Innovation Solution
A personal authentication apparatus and method that divides eyeball images based on brightness information, performs learning processing on these regions, and uses CNN weight information for accurate authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If iris authentication is performed under external light containing near-infrared components, then the authentication system can be used outdoors, but the authentication accuracy decreases due to exposure and reflection on the eyeball
Solution Approach 1:
The eyeball image is divided into multiple regions based on brightness information. Regions affected by external light exposure and reflection are segmented from regions with normal brightness, allowing different processing strategies to be applied to each region. This segmentation enables the system to handle outdoor conditions while maintaining authentication accuracy by focusing on unaffected regions.
Solution Approach 2:
Different learning processing is applied to different regions of the eyeball image based on their brightness characteristics. Regions with normal brightness undergo one type of learning processing, while regions affected by exposure or reflection undergo different processing. This local differentiation allows the system to optimize authentication accuracy for each region's specific condition.
2Measurement precision
If brightness of iris image is changed to match registered iris data conditions, then authentication accuracy improves under varying illuminance, but regions with different brightness levels due to external light exposure cannot be properly handled
Solution Approach 1:
Instead of uniformly changing the brightness of the entire iris image, the system segments the image into regions with different brightness characteristics. This allows selective processing where only affected regions are adjusted or excluded, while normal regions maintain their original characteristics for accurate authentication.
Solution Approach 2:
The system changes the approach from uniform brightness adjustment to region-specific parameter handling. By identifying regions affected by external light and applying different learning processing parameters to these regions versus normal regions, the system adapts to varying brightness conditions without compromising overall authentication accuracy.
Data Source
AI summary
A personal authentication apparatus comprises a first image capture unit that captures an eyeball image, a first image processing unit that obtains brightness information of the captured eyeball image and generates divided eyeball images by dividing a region of the eyeball image based on the brightness information of the eyeball image, a first storage unit that stores brightness information of the divided eyeball images and position information of the divided eyeball images in relation to the eyeball image in association with each other, and a learning unit that executes different types of learning processing respectively for the divided eyeball images in accordance with the brightness information and the position information of the divided eyeball images.