Eye Region Luminance Adjustment for Accurate Video Eye-State Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing devices struggle to analyze the state of an object, such as the degree of eye opening, as they do not adequately adjust luminance in regions including the eyelid, which is necessary for accurate eye part detection and analysis.
Innovation Solution
A video processing device that adjusts luminance values of both the iris and eyelid parts to enhance contrast, allowing for better recognition of eye contours and analysis, including the degree of eye opening, by comparing luminance differences and applying specific threshold adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If luminance adjustment is performed only on the iris part, then the face image has no incongruity in impression, but the eye part analysis (such as degree of opening) cannot be accurately performed
Solution Approach 1:
The patent segments the eye region into multiple parts (iris part and eyelid part) and applies different luminance adjustment strategies to each segment. The iris part luminance is adjusted to maintain natural appearance, while the eyelid part luminance is adjusted to enhance detectability for eye state analysis.
Solution Approach 2:
Different luminance adjustment characteristics are applied to different regions of the eye. The iris region receives adjustment to maintain natural impression, while the eyelid region receives adjustment optimized for detection accuracy, creating local quality differences in the image processing.
2Measurement precision
If luminance adjustment is performed on the eyelid part to enable eye analysis, then eye state detection becomes possible, but the naturalness of the face image may be compromised
Solution Approach 1:
The processing is segmented into iris region adjustment and eyelid region adjustment, allowing each region to be optimized for its specific purpose while maintaining overall image quality.
Solution Approach 2:
The patent changes luminance parameters differently for iris and eyelid regions based on their respective requirements, using parameter adjustments tailored to each region's function in the image.
3Illumination intensity
If high luminance is used to capture the face, then the face is clearly visible, but the iris part becomes too bright for proper analysis
Solution Approach 1:
The patent applies local quality adjustment by treating the iris region differently from the rest of the face. While the overall face maintains high luminance for visibility, the iris region receives specific luminance adjustment to bring it into the appropriate range for analysis.
Solution Approach 2:
Luminance parameters are changed selectively in the iris region to correct the over-brightness issue while preserving the overall high luminance level that ensures face visibility.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A video processing device (30) includes a video acquirer (31), a video recognition unit (32), a luminance value acquirer (33), and a luminance adjuster (35). The video acquirer (31) acquires a video in which the face of an image capture object is captured. The video recognition unit (32) recognizes an iris part and an eyelid part in a video acquired by the video acquirer (31). The luminance value acquirer (33) acquires a luminance value of each of the iris part and the eyelid part recognized by the video recognition unit (32). The luminance adjuster (35) adjusts a luminance value of a video such that the luminance values of the iris part and the eyelid part acquired by the luminance value acquirer (33) respectively become closer to a certain iris luminance upper threshold in the iris part and a certain eyelid luminance lower threshold in the eyelid part.