Frequency-Band Image Processing to Preserve Skin Texture

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional skin beautification techniques lose the natural texture of a person's skin while smoothing wrinkles and blemishes, resulting in an unnatural impression.

Innovation Solution

An image processing method that extracts and adjusts alternating current (AC) components of skin images to retain skin texture while reducing wrinkles and blemishes, using frequency-based amplitude adjustments and face detection for personalized correction.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If conventional skin beautification techniques (ε-filter) are applied to smooth wrinkles and blemishes, then skin defects are reduced, but skin texture is lost resulting in unnatural appearance

Engineering Contradiction:

Improveskin defect removalVSAvoidskin texture

Core Design Contradiction:

Manufacturing precisionVSLoss of information

Solution Approach 1:

The patent segments skin image information into multiple frequency components using wavelet transform. Different frequency bands correspond to different skin features: high-frequency components capture skin texture details, while low-frequency components represent overall skin tone and large defects. This segmentation allows selective processing of different frequency components to remove defects while preserving texture.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies different processing strategies to different frequency components locally. High-frequency components undergo selective filtering to preserve texture while reducing noise, medium-frequency components are processed to remove blemishes, and low-frequency components are smoothed to reduce wrinkles. This local quality approach ensures that each frequency band is treated appropriately to maintain natural skin appearance.

Inventive Principle:

Principle #3Local quality

2Manufacturing precision

If strong filtering is applied to remove blemishes and wrinkles, then skin appearance is improved, but fine irregularities and texture are also removed

Engineering Contradiction:

Improveblemish and wrinkle removalVSAvoidskin surface irregularities

Core Design Contradiction:

Manufacturing precisionVSShape

Solution Approach 1:

The patent employs dynamic thresholding and adaptive filtering parameters based on local image characteristics. The filtering strength is adjusted dynamically according to the frequency component being processed and local skin features, allowing stronger filtering for defects while maintaining gentler processing for texture preservation in different regions.

Inventive Principle:

Principle #15Dynamics

Solution Approach 2:

The patent changes processing parameters across different frequency bands. Wavelet decomposition levels, filtering thresholds, and reconstruction weights are adjusted for each frequency component. High-frequency components use parameters that preserve fine details, while low-frequency components use parameters that emphasize defect removal, achieving selective smoothing without texture loss.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentEP4610926A1Image processing apparatus and method, electronic apparatus, program, and storage medium

Publication Date: 2025.09.03 CANON KK

Figure 1
Figure 2~3
Figure 4

AI summary

An image processing apparatus comprises: input means configured to input an image signal; extracting means configured to extract a first signal in a predetermined first frequency band from the image signal; and adjusting means configured to adjust an amplitude of the first signal so that the amplitude falls within a predetermined amplitude range.