3D Face Shape Derivation Using Preset Feature Points

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for estimating a three-dimensional shape of a subject's face from multiple image frames face challenges in accurately mapping facial feature points due to potential one-to-many or many-to-one relationships between image frames, complicating the processing and introducing errors through outlier removal in optical flow calculations.

Innovation Solution

Presetting a predetermined number of facial feature points (e.g., 10 points) and acquiring face information for each image frame to uniquely identify their positions, allowing for straightforward and accurate mapping without complex processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If optical flow calculation is used to map pixels between adjacent image frames, then mapping can be performed, but outlier removal processing is required to eliminate mapping errors, complicating the processing

Engineering Contradiction:
Improvemapping accuracyVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by presetting a predetermined number of facial feature points before image frame processing. This allows the system to directly acquire face information for these predetermined points without requiring complex optical flow calculations or outlier removal, thereby simplifying processing while maintaining mapping accuracy.

Inventive Principle:
Principle #10Preliminary action

2Measurement precision

If rectangular region matching is used to track feature points in time-series images, then feature points can be mapped, but one-to-many or many-to-one relationships may occur, complicating the mapping process

Engineering Contradiction:
Improvefeature point mapping accuracyVSAvoidmapping processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent resolves this contradiction by presetting a predetermined number of facial feature points with fixed positions. This preliminary action ensures that the same number of feature points are extracted from each image frame, eliminating one-to-many or many-to-one mapping relationships and simplifying the mapping process while maintaining accuracy.

Inventive Principle:
Principle #10Preliminary action

3Adaptability or versatility

If the number of feature points is not fixed in one image frame, then feature detection is flexible, but one-to-one mapping between feature points in different image frames cannot be guaranteed

Engineering Contradiction:
Improvefeature detection flexibilityVSAvoidone-to-one mapping accuracy
Core Design Contradiction:
Adaptability or versatilityVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by presetting a predetermined number of facial feature points before processing. This ensures that the same number of feature points are extracted from each image frame, guaranteeing one-to-one mapping between frames while maintaining the flexibility of feature detection through the use of face information acquisition for each frame.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3699865B1Three-dimensional FACE shape derivation device, three-dimensional FACE shape deriving method, and non-transitory computer readable medium
Publication Date: 2022.11.30 NEC CORP
  • EP3699865B1 patent drawingFigure 1~2
  • EP3699865B1 patent drawingFigure 3
  • EP3699865B1 patent drawingFigure 4

AI summary

A three-dimensional facial shape estimating device (300) includes a face image acquiring unit (301) configured to acquire a plurality of image frames that capture a subject's face; a face information acquiring unit (302) having, preset therein, a predetermined number of facial feature points necessary for estimating a three-dimensional shape of the subject's face, the face information acquiring unit (302) being configured to acquire, for each of the plurality of image frames, face information that indicates a position of each of the predetermined number of facial feature points of the subject's face within the image frame; and a three-dimensional shape estimating unit (303) configured to perform mapping of each of the predetermined number of facial feature points of the subject's face between the plurality of image frames based on the face information of each of the plurality of image frames and to estimate the three-dimensional shape of the subject's face based on a result from the mapping.