Image Processing Device for Natural Makeup Simulation

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Conventional makeup simulation technologies require ideal, inexpressive front-facing images with even lighting, making it difficult to perform makeup simulations in natural or unprepared states, such as with glasses or messy hair, and often result in unnatural outcomes due to incorrect face orientation, object coverage, and uneven lighting.

Innovation Solution

An image-processing device and method that identifies the degree of skin color in each region of an image and colors the skin accordingly, allowing for proper makeup application even in non-ideal conditions by distinguishing between skin and non-skin areas, such as hair or glasses, and adjusting makeup placement and color depth.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If conventional makeup simulation technologies are used, then makeup can be applied to ideal front-facing images with even lighting, but the system cannot handle natural or unprepared states such as glasses or messy hair

Engineering Contradiction:

Improveadaptability to various image conditionsVSAvoidaccuracy of makeup application

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The patent segments the face image into multiple regions (forehead, eyes, cheeks, chin) and identifies tracking points within each region. This segmentation allows the system to independently analyze and process different facial areas, enabling reliable makeup application even when parts of the face are obscured by glasses or hair, as long as sufficient tracking points are visible in other regions.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent applies local quality by using skin color information from specific tracking points and their surrounding regions to determine makeup application parameters. Instead of treating the entire face uniformly, the system uses local skin color characteristics at each tracking point to guide makeup application, ensuring natural and accurate results adapted to local facial conditions.

Inventive Principle:

Principle #3Local quality

2Manufacturing precision

If ideal front-facing images with even lighting are required, then accurate makeup application can be achieved, but the user must visit cosmetic stores and prepare specifically for image capture

Engineering Contradiction:

Improveprecision of makeup simulationVSAvoidease of performing makeup simulation

Core Design Contradiction:

Manufacturing precisionVSEase of operation

Solution Approach 1:

The patent implements self-service by enabling users to perform makeup simulation independently at home using ordinary photographs. The system automatically detects facial features, identifies tracking points, and applies makeup without requiring professional assistance or special preparation, making the process as easy as taking a regular photo with a digital camera or smartphone.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The patent changes the parameter requirements for image capture by accepting ordinary photographs with varied lighting conditions, angles, and compositions. Instead of requiring specific lighting parameters and precise positioning, the system uses robust feature detection algorithms that can identify facial landmarks across a wide range of imaging parameters, thereby simplifying the ease of operation.

Inventive Principle:

Principle #35Parameter changes

3Ease of manufacture

If makeup is applied by superimposing color on face image, then makeup simulation can be performed, but unnatural outcomes occur due to incorrect face orientation or object coverage

Engineering Contradiction:

Improvesimplicity of makeup application processVSAvoidaccuracy of makeup placement

Core Design Contradiction:

Ease of manufactureVSManufacturing precision

Solution Approach 1:

The patent performs preliminary action by detecting facial features and identifying tracking points before applying makeup. The system analyzes the face image, locates key facial landmarks (eyes, cheeks, chin), and establishes a mapping between tracking points and makeup application regions. This preliminary detection and mapping ensures that subsequent makeup application is accurately positioned, even in non-ideal image conditions.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent uses feedback by continuously monitoring the quality and distribution of tracking points to adjust makeup application parameters. The system evaluates whether sufficient tracking points are detected and whether they are properly distributed across facial regions, using this feedback to determine whether to proceed with makeup application or to request a different image, thereby maintaining high precision in makeup placement.

Inventive Principle:

Principle #23Feedback

Data Source

PatentEP2615576B1Image-processing device, image-processing method, and control program

Publication Date: 2019.11.13 OMRON CORP

Figure 1
Figure 2~4
Figure 5~6

AI summary

The purpose of the present invention is to implement an image-processing device and image-processing method that can apply appropriate makeup to face images taken under a variety of conditions. This image-processing device (6), which performs a process in which makeup of a specified color is composited onto skin in a face image, is provided with: a skin-identification unit (22) that identifies the degree of skin color in the color of each pixel in at least part of the face image; and a compositing unit (17) that composites makeup onto the face image in intensities corresponding to said skin-color degrees.