Iris Detection Using Luminance Edge Chords in Face Regions
Solution Overview
Problem
Existing iris detection techniques in digital images face challenges with poor detection rates due to variations in brightness and iris color, and high processing requirements that make real-time detection unrealistic on embedded devices.
Innovation Solution
A method for robust iris detection using edge detection in a predetermined face region, with adaptive thresholding and overlapping windows to identify sclera edge pixels and chords, allowing for accurate iris identification under varying conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If existing iris detection techniques are used, then detection can be performed on digital images, but detection rate is poor due to variations in brightness levels and iris color
Solution Approach 1:
The patent transforms the iris detection problem from color-based to luminance-based by changing the parameter domain. It converts RGB images to grayscale and applies adaptive thresholding, thereby eliminating the impact of color variations and improving detection reliability across different iris colors and lighting conditions
Solution Approach 2:
The patent employs dynamic adaptive thresholding where the threshold value is not fixed but adaptively determined from the image data itself using histogram analysis. This dynamic adjustment allows the detection algorithm to automatically adapt to varying brightness levels and lighting conditions in different images
2Productivity
If existing iris detection techniques are used, then iris detection can be performed, but processing requirements are high making real-time detection unrealistic on embedded devices
Solution Approach 1:
The patent extracts only the essential luminance information from the original color image by converting to grayscale, discarding color data that is not necessary for iris detection. This extraction reduces the data processing load while maintaining detection effectiveness, enabling real-time processing on embedded devices
Solution Approach 2:
The patent segments the face region into multiple zones and applies edge detection only in specific predetermined regions where the iris is likely to be located. This spatial segmentation reduces the overall processing area and computational requirements compared to analyzing the entire image
Data Source
AI summary
A system and a method for iris detection in a face region of a digital image are disclosed. In one embodiment, iris is detected in the face region of the digital image using eye features, such as a shape of the iris and/or luminance values of a sclera region.