Digital Image Color Correction Using Sclera and Pupil Reference
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image processing techniques for correcting skin color fail to accurately account for facial skin color and lighting environment, often requiring specialized devices and complex preparations, and struggle with extracting correct skin tones due to wide color ranges including shadows and highlights.
Innovation Solution
A method and system that detect the face and eyes from an image, extract the sclera and pupil using the Otsu algorithm and eye-shape mask, compare these areas to reference values stored in a database, and correct the image by converting colors based on brightness thresholds, allowing for accurate extraction of facial skin color regardless of lighting conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional skin color measurement devices are used, then accurate skin color measurement is achieved, but device complexity and cost increase
Solution Approach 1:
The system uses the human eye's sclera and pupil as natural reference objects for color calibration, eliminating the need for external specialized measurement devices. The eye itself provides the reference white point (sclera) and dark point (pupil) needed for accurate skin color measurement under various lighting conditions.
Solution Approach 2:
The sclera and pupil act as intermediary reference objects between the light source and the skin being measured. By comparing skin tones against these known reference points in the image, the system can calculate and correct for lighting effects without requiring direct spectral measurement equipment.
2Ease of operation
If standard color table preparation is used, then lighting environment measurement is simplified, but operational complexity increases due to preparation requirements
Solution Approach 1:
The system automatically identifies and uses the sclera and pupil in the captured image as reference standards, eliminating the need for manual preparation of color tables or calibration charts. The reference information is extracted directly from the subject's own anatomy in real-time.
Solution Approach 2:
The system pre-identifies the sclera and pupil regions in the captured image before performing skin color analysis. By detecting eye location and segmenting the sclera and pupil areas in advance, the reference values are ready for immediate use in lighting correction without requiring separate preparation steps.
3Quantity of substance
If wide skin color range extraction is used, then more skin areas are captured, but measurement precision decreases due to inclusion of shadows and highlights
Solution Approach 1:
The system transforms the skin color data from RGB space to L*a*b* color space, where the L* channel represents lightness and the a* and b* channels represent color information. By analyzing the distribution of L* values and identifying the dominant color range, the system can filter out extreme values corresponding to shadows and highlights while retaining representative skin tones.
Solution Approach 2:
The system applies different processing strategies to different regions within the extracted skin area. By detecting the distribution characteristics of skin tones and identifying regions with extreme brightness values, it selectively weights or excludes certain areas based on their local properties, giving more importance to mid-tone regions that represent true skin color.
Data Source
AI summary
The color correction method in accordance with an embodiment of the present invention can comprise the steps of: acquiring an image from a captured image; detecting the face and eyes from the captured image; separating the sclera and pupil of the detected eye(s); correcting the image by comparing the areas of the sclera and pupil extracted by separating the sclera and pupil with the reference values stored in a database; and extracting the skin color of the face from the corrected image.