Eye Baseline Alignment for Gaze Estimation Efficiency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In gaze estimation technology, appearance-based approaches face challenges in efficiently analyzing eye movement information due to high-dimensional image spaces and the need for high resolution to capture eye details, which reduces efficiency and accuracy.
Innovation Solution
An image processing method that determines vertical and horizontal baselines corresponding to the eye boundary in an input image, aligns the image based on these baselines, and extracts movement responsivity to predict gaze direction, using projection techniques and skin modeling to enhance alignment and data reduction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high resolution is used to capture eye details such as eye lids and eye lashes, then the accuracy of eye feature capture is improved, but the efficiency of gaze applications is reduced
Solution Approach 1:
The patent extracts only the essential eye features (pupil, iris, sclera boundaries) needed for gaze estimation while discarding unnecessary high-resolution details such as eye lids and eye lashes. This is achieved through image processing techniques that identify and isolate critical eye regions, thereby maintaining measurement precision for gaze-related features while reducing overall data volume to improve processing efficiency.
Solution Approach 2:
The patent applies different processing quality levels to different regions of the eye image. Critical regions such as the pupil and iris boundaries are processed with high precision to maintain accurate gaze estimation, while non-critical regions are processed at lower resolution or aggregated. This local differentiation allows the system to maintain necessary measurement precision while reducing overall computational load.
2Loss of information
If the entire eye image including detailed features is processed, then comprehensive eye information is captured, but the complexity of data processing increases
Solution Approach 1:
The patent segments the eye image into distinct functional regions (pupil, iris, sclera, and boundary regions) and processes each segment separately using appropriate algorithms. This segmentation allows the system to capture comprehensive eye information by systematically analyzing each region while reducing overall processing complexity through divide-and-conquer strategies, where each segment can be handled with optimized, simpler processing techniques.
Solution Approach 2:
The patent extracts only the essential eye features (pupil, iris, sclera boundaries) needed for gaze estimation while discarding unnecessary high-resolution details such as eye lids and eye lashes. This is achieved through image processing techniques that identify and isolate critical eye regions, thereby maintaining measurement precision for gaze-related features while reducing overall data volume to improve processing efficiency.
3Loss of information
If high-dimensional image space is used for appearance-based gaze analysis, then comprehensive eye appearance information is retained, but the computational efficiency is reduced
Solution Approach 1:
The patent transforms the high-dimensional image space data into a lower-dimensional feature space by extracting key geometric and appearance parameters from segmented eye regions. Instead of processing the entire high-dimensional pixel matrix, the system converts image data into essential dimensional representations (such as pupil center coordinates, iris boundaries, and relative positions) that preserve the necessary appearance information for gaze analysis while dramatically reducing computational dimensions and improving efficiency.
Data Source
AI summary
An image processing method includes determining a vertical baseline and a horizontal baseline corresponding to a boundary of an eye of a user in an input image of the eye of the user, and aligning the input image based on the vertical baseline and the horizontal baseline.