Gaze Position Estimation via Eye Area Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gaze position estimation systems require heavy processing to detect the center point of the user's iris and estimate the projection position of the eyeball, leading to high processing loads and the need for high-resolution cameras, which limits precision and increases computational burden.
Innovation Solution
A system that estimates the gaze position by detecting the eye area, face area, and other facial features without requiring iris detection or eyeball projection estimation, using areal size and position information to calculate the gaze point on a screen based on differential values and moving averages.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If iris center detection and eyeball projection estimation are performed, then gaze position estimation accuracy is improved, but processing load increases significantly
Solution Approach 1:
The patent extracts only the necessary information (eye area position and size) from the full eye image, eliminating the need for complex iris center detection and eyeball projection estimation. This selective extraction maintains sufficient accuracy for gaze position estimation while dramatically reducing computational complexity and processing load.
2Measurement precision
If high-resolution camera is used to capture detailed eye features, then iris detection precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent changes the measurement parameters from detailed iris features to coarse eye area position and size. This parameter transformation allows the use of lower-resolution cameras while maintaining adequate measurement precision for gaze estimation, thereby reducing device complexity and cost.
3Measurement precision
If complex image processing algorithms are applied for iris and eyeball analysis, then gaze position accuracy is improved, but processing time increases
Solution Approach 1:
The patent applies partial action by performing only the minimum necessary processing (detecting eye area position and size) without executing the full sequence of complex algorithms (iris detection, eyeball projection estimation). This selective processing achieves adequate gaze position accuracy while significantly reducing processing time.
Data Source
AI summary
A photographing unit photographs a face of the user who is looking at a screen displayed on a display unit. An area detecting unit detects, from the photographed image of the photographing unit, an eye area of the user and at least one of a face area of the user or a predetermined part area of the user other than the user's eyes. An areal size/position information obtaining unit obtains areal size information and position information of the eye area, and areal size information and position information of the at least one of the face area or the predetermined part area. A gaze position estimation unit estimates a position in the screen that the user is gazing at, based on the areal size information and the position information.