Facial Image Perspective Correction for Standardized Recognition

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing facial image capture technologies face challenges in ensuring standardized and high-quality images for automated facial recognition, particularly in limited spaces and with non-standardized distances, leading to higher classification errors and reduced user-friendliness and throughput in security-critical applications.

Innovation Solution

A device and method that corrects facial images by performing a three-dimensional image transformation to align the captured image with a predefined spatial position, using a spatial orientation determination device to determine the initial distance and orientation, and applying perspective distortion to achieve a standardized facial image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If facial images are captured from any distance and perspective in limited spaces, then ease of operation and adaptability are improved, but manufacturing precision and measurement precision deteriorate due to non-standardized distances and perspectives

Engineering Contradiction:
Improveease of operationVSAvoidstandardization of facial image
Core Design Contradiction:
Ease of operationVSManufacturing precision

Solution Approach 1:

The system changes the spatial parameters (distance, orientation, perspective) of the captured facial image by applying a three-dimensional image transformation. The transformation maps pixels from the first spatial position to a second spatial position, effectively adjusting the image parameters to meet standardization requirements while allowing flexible capture conditions.

Inventive Principle:
Principle #35Parameter changes

Solution Approach 2:

The patent transitions from a two-dimensional image plane to a three-dimensional spatial transformation. By determining the first spatial position (including distance and orientation) and applying a 3D transformation, the system corrects perspective distortion and produces a standardized facial image that appears as if captured from a predefined second spatial position.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Adaptability or versatility

If facial images are captured from any distance and perspective, then adaptability is improved, but classification accuracy deteriorates due to insufficient image quality

Engineering Contradiction:
ImproveadaptabilityVSAvoidclassification accuracy
Core Design Contradiction:
Adaptability or versatilityVSReliability

Solution Approach 1:

The system applies a three-dimensional image transformation that adjusts spatial parameters (distance, orientation, perspective) to produce a standardized facial image. This transformation corrects perspective distortion and ensures the image meets quality requirements for automated facial recognition, thereby maintaining high classification accuracy while allowing flexible capture conditions.

Inventive Principle:
Principle #35Parameter changes

3Manufacturing precision

If standardized facial images are captured at mandatory minimum distances, then manufacturing precision and measurement precision are improved, but device complexity and ease of operation worsen due to space requirements

Engineering Contradiction:
Improvestandardization of facial imageVSAvoiddevice complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent replaces the mechanical approach of enforcing minimum capture distances with a computational three-dimensional image transformation. Instead of requiring physical space constraints, the system uses software-based spatial transformation to achieve standardization, thereby reducing device complexity and eliminating the need for mandatory minimum distance requirements.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP3646236B1Apparatus, method and computer program for correcting a facial image of a person
Publication Date: 2025.12.24 BUNDESDRUCKEREI GMBH
  • EP3646236B1 patent drawingFigure 1
  • EP3646236B1 patent drawingFigure 2
  • EP3646236B1 patent drawingFigure 3

AI summary

The invention relates to an apparatus (100) for correcting a facial image of a person. The apparatus (100) comprises a photography camera (101), which is embodied to capture an image of a face of the person in order to obtain the facial image, wherein the photography camera (101) is arranged in a first spatial position relative to the face of the person. The apparatus (100) further comprises a spatial position determination device (103), which is embodied to determine the first spatial position using a predetermined reference facial model, wherein the predetermined reference facial model represents a predetermined reference facial form. Moreover, the apparatus (100) comprises a processor (105), which is embodied to determine a spatial position deviation of a second spatial position from the first spatial position and distort the facial image in perspective fashion on the basis of the spatial position deviation in order to obtain a corrected facial image. Further, the apparatus (100) may comprise a communications interface (107), which is embodied to transmit the corrected facial image to a server.