Eye Region Area Comparison for Asymmetric Eye State Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems cannot accurately detect the motion of closing only one eye from a facial image, as they primarily focus on detecting open or closed eyes without differentiating between the two eyes independently.
Innovation Solution
A method for determining whether one eye is closed by calculating and comparing the eye region areas of the right and left eyes from a facial image, using a ratio between the eye region values and setting search regions based on feature points to accurately identify the eye regions, allowing for efficient processing and robustness against individual differences and photographing conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional eye detection methods are used, then open or closed eye state can be detected, but the motion of closing only one eye cannot be detected
Solution Approach 1:
The invention divides the eye detection task into separate processing for the right eye and left eye. By calculating eye region areas independently for each eye and comparing them, the system can detect asymmetric eye closing motions that conventional unified detection methods miss.
2Measurement precision
If detailed image processing is performed to accurately detect eye regions, then detection precision improves, but processing complexity increases
Solution Approach 1:
The invention extracts only the essential feature needed for detection - the eye region area - from the facial image. By focusing on this specific extracted feature and comparing areas between eyes, the system achieves accurate detection without requiring complex processing of the entire image.
Solution Approach 2:
The invention changes the detection parameter from analyzing detailed eye structure or multiple image features to simply comparing eye region areas. This parameter simplification reduces processing complexity while maintaining detection accuracy for one eye closing motions.
3Device complexity
If eye detection is performed without considering individual differences, then processing is simpler, but detection reliability decreases
Solution Approach 1:
The invention leverages the asymmetric change in eye region areas when one eye closes, rather than trying to account for all individual differences. By detecting the relative area difference between eyes rather than absolute measurements, the system achieves reliable detection across different users without complex individual calibration.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A game apparatus includes a camera cartridge, and detects a plurality of feature points from a facial image of a user photographed by the camera cartridge. The game apparatus calculates areas of black eye regions of a left eye and a right eye by utilizing the plurality of feature points. Then, the game apparatus detects whether or not any one of the left eye and the right eye is closed on the basis of the calculated areas. The game apparatus executes game processing according to the detection result.