Virtual Hair Coloring Using Frequency Histogram Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current virtual hair coloring techniques fail to accurately preserve the natural details and variations of hair colors, leading to unrealistic results, and are not capable of real-time processing on mobile devices, which limits their effectiveness in providing a realistic and dynamic virtual try-on experience.
Innovation Solution
A computer-implemented method and system that maps gray levels from a swatch image to a hair portion of an input image using deep neural networks and GPUs, employing a mapping table to determine swatch color values for each pixel, allowing for real-time recoloring and preserving the natural texture and brightness of hair.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional virtual hair coloring methods are used, then processing speed is improved, but realism and detail preservation deteriorate
Solution Approach 1:
The patent segments the hair coloring process into multiple independent components: gray level extraction, frequency histogram calculation, mapping table generation, and pixel-level color application. This segmentation allows each component to be optimized independently, enabling real-time processing while preserving detailed hair characteristics through dedicated frequency-matching operations.
Solution Approach 2:
The patent transforms the hair coloring problem from direct color replacement to a parameter-based transformation system. By converting hair pixels to gray levels, calculating frequency distributions, and mapping to target color parameters, the system maintains detailed variations while achieving realistic coloring results at real-time speeds.
2Manufacturing precision
If complex algorithms are used to preserve hair details, then realism is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary actions by pre-calculating frequency histograms from the target swatch image and pre-generating mapping tables that define the relationship between gray levels and target colors. This preliminary processing enables the actual hair coloring to be performed rapidly by simply looking up and applying pre-determined color mappings, achieving both realism and speed.
Solution Approach 2:
The patent replaces complex iterative mechanical coloring algorithms with a direct lookup-based system. Instead of repeatedly adjusting and rendering colors, the system substitutes this with efficient frequency-matching and table-lookup operations, dramatically reducing processing time while maintaining detailed realism through frequency-based color assignment.
3Measurement precision
If frequency matching is used to preserve color variations, then color accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent introduces frequency histograms and mapping tables as intermediary data structures that bridge the source hair image and target color swatch. These intermediaries simplify the computational process by pre-organizing color relationships, allowing accurate color matching without requiring complex real-time calculations during the actual coloring operation.
Data Source
AI summary
There is provided methods and systems for a hair virtual try on. Gray levels from a swatch image and gray levels from a hair portion of an input image are mapped by matching their respective frequencies to establish a map relationship between the swatch image and the hair portion, wherein the gray levels of the swatch image are associated to respective swatch color values. A pixel in the hair portion is colored based on a swatch color value determined using a gray level of the pixel and the map relationship.