Iris Tracking via Luminance Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Eye tracking technologies face challenges in accurately tracking subtle eye movements due to variations in object appearance, such as size, shape, and lighting conditions, especially in low-quality images, making it difficult to identify and follow the iris effectively.
Innovation Solution
An eye tracking method that estimates the expected size and location of the iris by applying a face detection algorithm and comparing luminance values within a region, using template matching and non-linear manipulation of pixel values to identify the darkest region representing the iris, which is then tracked across frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional eye tracking methods are used to accurately track subtle eye movements, then tracking precision is improved, but processing resources and computational complexity increase significantly
Solution Approach 1:
The patent divides the face image into multiple regions of interest (eyes, eyebrows, nose, mouth) and processes only these segmented areas rather than the entire image. This segmentation allows the system to focus computational resources on eye tracking while reducing overall processing complexity and resource requirements.
Solution Approach 2:
The patent applies different processing strategies to different facial regions. Specifically, it uses luminance-based detection for iris identification within predefined eye regions, rather than applying complex algorithms to the entire face. This local quality approach optimizes processing efficiency while maintaining tracking precision in critical areas.
2Measurement precision
If complex algorithms are applied to identify iris location and size, then tracking accuracy is improved, but processing time and computational load increase
Solution Approach 1:
The patent performs preliminary actions by pre-defining eye regions based on face geometry and estimating iris location and size before detailed analysis. This preliminary positioning narrows down the search area, allowing subsequent iris identification to be performed quickly using simple luminance comparisons rather than exhaustive complex algorithms.
Solution Approach 2:
The patent uses a simplified luminance-based detection method within predefined regions rather than applying full complex image processing algorithms. This partial action approach achieves sufficient tracking accuracy for eye movements while significantly reducing processing time and computational load.
3Reliability
If the entire face image is processed to locate eyes, then detection reliability is improved, but processing efficiency decreases
Solution Approach 1:
The patent segments the face image into distinct regions (eyes, eyebrows, nose, mouth) using geometric relationships and facial feature detection. By processing only the eye regions rather than the entire face, the system maintains detection reliability through proper region identification while significantly improving processing efficiency.
Solution Approach 2:
The patent applies specialized processing only to local eye regions rather than the entire face image. This local quality approach uses luminance-based detection within predefined eye areas, ensuring reliable eye detection while maximizing processing efficiency by avoiding unnecessary computation in other facial regions.
Data Source
AI summary
An eye tracking method includes: in a frame of a series of acquired frames, estimating an expected size and expected location of an image of an iris of an eye within that frame; and determining a location of the iris image within that frame by identifying a region within the expected location, the size of the region being consistent with the expected size, pixels of that region being darker than pixels of other such regions within that frame.