Face Mask Generation for Selective Image Processing

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Users of digital images struggle to effectively adjust specific portions of images, such as enhancing or de-emphasizing facial features, without affecting other parts of the image, due to the lack of efficient image processing techniques that can accurately identify and isolate facial regions.

Innovation Solution

The method involves identifying face regions in images using facial landmarks, determining associated face masks to isolate skin pixels, and applying processing operations like blurring or contrast enhancement selectively to these pixels while excluding non-face pixels, using techniques like blobbing and color thresholding to define spatial face areas and falloff regions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If image processing operations are applied to entire images, then processing coverage is complete, but facial features cannot be selectively adjusted without affecting other areas

Engineering Contradiction:

Improveselective adjustment capabilityVSAvoidimage processing complexity

Core Design Contradiction:

Ease of operationVSDevice complexity

Solution Approach 1:

The image is segmented into face regions and non-face regions using facial landmark detection. The system identifies key facial points (eyes, nose, mouth) and uses these landmarks to define a face mask that separates facial pixels from non-facial pixels, enabling selective processing of different image regions

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

A face mask serves as an intermediary data structure between the original image and the processing operation. The mask contains pixel values that indicate whether each pixel belongs to a face region, allowing processing operations to selectively apply to only the desired regions based on the mask guidance

Inventive Principle:

Principle #24Intermediary (Mediator)

2Manufacturing precision

If processing operations are applied to enhance facial features, then facial quality is improved, but non-face areas may be degraded or over-processed

Engineering Contradiction:

Improvefacial feature accuracyVSAvoidnon-face area degradation

Core Design Contradiction:

Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

Different processing operations are applied to different regions of the image based on the face mask. Facial regions receive skin smoothing and feature enhancement operations, while non-facial regions are excluded from processing or receive different treatment, ensuring each region is processed appropriately for its content

Inventive Principle:

Principle #3Local quality

Solution Approach 2:

Instead of trying to protect face regions from processing, the system inverts the approach by creating a mask that explicitly identifies face regions and then applies processing only to those masked regions. The inverse mask (non-face regions) is used to exclude areas that should not be processed

Inventive Principle:

Principle #13The other way round (Inversion)

3Ease of operation

If manual masking is used to isolate facial regions, then processing selectivity is achieved, but user effort and time are significantly increased

Engineering Contradiction:

Improveprocessing selectivityVSAvoidmask creation time

Core Design Contradiction:

Ease of operationVSLoss of time

Solution Approach 1:

The system performs automatic facial landmark detection and face mask generation without requiring manual user input. The algorithm independently identifies facial features, determines face regions, and creates the processing mask automatically, eliminating the need for users to manually create masks while still achieving precise selective processing

Inventive Principle:

Principle #25Self-service

4Productivity

If simple color thresholding is used to identify skin pixels, then processing speed is improved, but accuracy in distinguishing skin from non-skin pixels deteriorates

Engineering Contradiction:

Improveprocessing speedVSAvoidskin pixel identification accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary facial landmark detection to identify the approximate location and boundaries of the face before applying color-based pixel classification. This preliminary spatial constraint narrows the search area and provides context that improves the accuracy of subsequent color thresholding operations

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system combines multiple approaches: facial landmark-based spatial region definition is merged with color thresholding for pixel classification. By integrating spatial constraints from landmarks with color information, the system achieves both speed and accuracy in identifying skin pixels

Inventive Principle:

Principle #5Merging (Combining)

Data Source

PatentUS8983152B2Image masks for face-related selection and processing in images

Publication Date: 2015.03.17 GOOGLE LLC

AI summary

In some implementations, a method includes identifying one or more face regions of an image, the face regions including pixels that depict at least a portion of one or more faces of persons. The face regions are identified based on identifying facial landmarks of the faces. The method determines an associated face mask for each of the faces based on the face regions, where each face mask indicates which pixels in the image depict the corresponding face. Face pixels can be selected for processing by applying the face masks, and image pixels outside the faces can be selected by inversely applying the face masks. The selected pixels can be provided to a processing operation for adjustment of the selected pixels.