Skin Color Detection Using YUV Chrominance Lookup Matrix

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing skin color detection methods face challenges under varying illumination conditions, leading to inaccurate and non-contiguous detection results due to high calculation loads and pixel value fluctuations, which affect the accuracy of skin color detection in dynamic images.

Innovation Solution

A skin color detection method that generates a skin color index matrix by processing images under various illumination conditions, using chrominance signal values in the YUV domain to determine the probability of pixel points being skin color points, thereby improving detection accuracy and consistency across different lighting conditions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the Bayes formula is used to calculate skin color probability, then the detection accuracy is improved, but the calculation load increases significantly

Engineering Contradiction:
Improveskin color detection accuracyVSAvoidcalculation load
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent pre-calculates and stores skin color probability values for all possible chrominance combinations in a lookup table during system initialization. This preliminary action eliminates the need for complex Bayes formula calculations during actual skin color detection, reducing real-time computational load while maintaining detection accuracy.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a simplified copy of the probability calculation results in the form of a lookup table, storing pre-computed skin color probabilities for different chrominance values. This copy allows rapid retrieval during detection without performing the original complex calculations, effectively reducing computational complexity.

Inventive Principle:
Principle #26Copying

2Ease of operation

If a predetermined pixel value range is used for skin color detection, then the detection process is simplified, but the detection accuracy decreases under varying illumination conditions

Engineering Contradiction:
Improvedetection process simplicityVSAvoidskin color detection accuracy
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent changes the detection parameter from fixed RGB pixel value ranges to chrominance values (U and V components) in the YUV color space. This parameter change makes the detection more robust to illumination variations because chrominance values are less sensitive to lighting changes than luminance or RGB values, thereby improving accuracy while maintaining simple threshold-based detection.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent pre-calculates skin color probability values for different chrominance combinations and stores them in a lookup table. This preliminary computation allows the system to use simple table lookup during detection rather than complex calculations, maintaining operational simplicity while achieving high accuracy across varying illumination conditions.

Inventive Principle:
Principle #10Preliminary action

3Device complexity

If fixed detection thresholds are used, then the detection algorithm is simple, but the detection results are non-contiguous and inconsistent under different illumination conditions

Engineering Contradiction:
Improvedetection algorithm complexityVSAvoiddetection result consistency
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent transitions from using fixed RGB threshold ranges to using chrominance-based probability lookup. This parameter change enables the system to handle illumination variations more effectively by relying on chrominance values that are more stable under different lighting conditions, improving result consistency while keeping the algorithm relatively simple through table lookup.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent introduces dynamic adaptability by using a lookup table that contains probability values for various chrominance combinations. This allows the detection threshold to effectively adapt to different illumination conditions without changing the algorithm structure, maintaining simplicity while achieving consistent and contiguous detection results across varying environments.

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3678056B1Skin color detection method and device and storage medium
Publication Date: 2023.09.06 BIGO TECH PTE LTD
  • EP3678056B1 patent drawingFigure 1~2
  • EP3678056B1 patent drawingFigure 3A~3B
  • EP3678056B1 patent drawingFigure 3C~3D

AI summary

The present invention relates to the technical field of image processing, and disclosed thereby are a skin color detection method and device and a storage medium. The method comprises: determining, for each pixel point in a target image to be subjected to skin color detection, a chrominance signal value of a pixel value corresponding to the pixel point in a YUV region (101); searching for a skin color probability value corresponding to the chrominance signal value in a stored skin color index matrix on the basis of the chrominance signal value of the pixel value corresponding to the pixel point in the YUV region (102), wherein the skin color index matrix is generated and obtained after processing skin color images under various lighting conditions, and the skin color probability value is the probability that the pixel point is a skin color point; and carrying out skin color detection on t he basis of the skin color probability value corresponding to the chrominance signal value of each pixel point in the target image (103). With the present invention, the probability that the pixel point in the target images is a skin color point is determined by means of the skin color index matrix, thereby carrying out skin color detection, and thus preventing the problem of detection result inaccuracy due to different lighting conditions, and improving skin color detection accuracy.