Portrait Skin Color Mapping for Diverse Tone Beautification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional skin color adjustment methods in image beautification are inflexible and inaccurate, failing to adapt to the diverse skin tones of different races and individuals, leading to suboptimal beauty effects.
Innovation Solution
A method and apparatus that utilize multiple skin color mapping tables corresponding to predetermined categories, determined by neural networks, to calculate weighted pixel values for flexible and accurate skin color adjustments in portraits, considering various skin tone probabilities and preferences.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If a fixed whitening treatment is applied to all skin colors, then the processing simplicity is improved, but the adaptability to different skin types deteriorates
Solution Approach 1:
The patent segments the continuous skin color spectrum into multiple discrete categories (e.g., 5 categories from light to dark). Each category has its own dedicated mapping table optimized for that skin type. This segmentation allows the system to provide specialized treatment for each skin type while maintaining overall system simplicity through categorical classification.
Solution Approach 2:
The patent changes the parameter of skin color mapping by using different mapping tables for different skin color categories. Instead of a single fixed mapping, the system selects and applies appropriate mapping tables based on the detected skin color category, thereby adapting the processing parameters to match the specific skin type being treated.
2Adaptability or versatility
If multiple skin color mapping tables are used for different skin color categories, then the adaptability to different skin types is improved, but the device complexity increases
Solution Approach 1:
The patent performs preliminary classification of the input image's skin color category before applying the appropriate mapping table. This preliminary action (categorization) is done once at the beginning, and then the corresponding pre-prepared mapping table is selected and applied. This approach manages complexity by organizing multiple mapping tables into a structured selection process rather than requiring complex real-time decision-making.
Solution Approach 2:
The patent creates a universal beautification system that handles multiple skin types through a common framework. The same overall processing pipeline and algorithm structure are used for all skin types; only the specific mapping table changes based on category. This multi-functionality allows one system to serve all skin types without requiring completely separate processing paths for each category.
3Productivity
If conventional fixed whitening treatment is used, then the processing speed is improved, but the manufacturing precision of skin color adjustment deteriorates
Solution Approach 1:
The patent applies local quality by using different mapping tables optimized for specific skin color ranges. Each mapping table contains precisely tuned adjustment parameters tailored to the characteristics of its target skin category. This ensures that each skin type receives the most appropriate adjustment, improving precision without requiring complex real-time analysis.
Solution Approach 2:
The system performs preliminary categorization of skin color types before adjustment, which enables the selection of the most appropriate mapping table in advance. This preliminary classification step is computationally efficient and allows the system to quickly identify the correct mapping table, maintaining high processing speed while enabling precise, customized adjustment for each skin type.
Data Source
AI summary
A method and an apparatus for adjusting skin colors of a portrait, an electronic device and a storage medium are provided. The method includes: acquiring an initial portrait image to be adjusted; determining multiple skin color category probabilities that the initial portrait image respectively belongs to multiple predetermined skin color categories; calculating a mapped pixel value of each pixel in the initial portrait image according to the multiple skin color category probabilities and multiple skin color mapping tables respectively corresponding to the multiple skin color categories; and determining a skin color beautified image based on the mapped pixel value of each pixel. With the method, the skin color adjustment can be flexibly and accurately adapted to various skin color conditions.