3D Head Model Generation for Distortion-Free Biometric ID Images

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for personalizing ID documents with biometric facial features face challenges in capturing distortion-free two-dimensional facial images suitable for reliable biometric recognition, due to issues like lighting, pose, and image quality, which affects the recognition rate and reliability.

Innovation Solution

A method involving the creation of three-dimensional head model data and associated texture data to generate distortion-free two-dimensional facial images for storage on ID documents, using techniques like morphing to correct image sections and ensure high-quality biometric facial features for recognition.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If two-dimensional facial images are captured using conventional image capture methods, then the process is simple and cost-effective, but the images suffer from distortions and quality issues that reduce biometric recognition reliability

Engineering Contradiction:
Improvebiometric recognition reliabilityVSAvoidimage capture process complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent transitions from two-dimensional image capture to three-dimensional head model generation. By capturing multiple images from different angles and constructing a 3D head model, the system eliminates perspective distortions and lighting issues inherent in 2D images. The 3D model can then be rendered into idealized 2D facial images for biometric recognition, achieving both simplicity and high reliability.

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

Solution Approach 2:

The patent creates a digital copy of the person's head in three-dimensional space through virtual photogrammetry. This 3D digital twin captures all facial geometric and textural information without physical constraints, allowing generation of multiple idealized 2D views from the same 3D model without requiring multiple physical photo sessions.

Inventive Principle:
Principle #26Copying

2Manufacturing precision

If strict image capture specifications are enforced to ensure distortion-free images, then image quality improves, but the process becomes more complex and costly

Engineering Contradiction:
Improveimage qualityVSAvoidcapture process complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

By moving to three-dimensional representation, the system can generate multiple idealized 2D images from a single 3D head model. The 3D model inherently contains all geometric and textural information, allowing computational generation of distortion-free 2D views without requiring strict physical capture conditions for each individual image.

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

Solution Approach 2:

The patent performs preliminary 3D head model construction during the capture process, storing the 3D geometry and texture data. This preliminary 3D representation serves as a reusable template that can generate multiple idealized 2D images later without requiring re-capture under strict specifications, thus simplifying the overall process.

Inventive Principle:
Principle #10Preliminary action

3Measurement precision

If conventional 2D image capture is used for ID document personalization, then the process is straightforward, but the facial features cannot be reliably extracted for biometric recognition

Engineering Contradiction:
Improvefacial feature extraction accuracyVSAvoidpersonalization process ease
Core Design Contradiction:
Measurement precisionVSEase of operation

Solution Approach 1:

The patent uses three-dimensional head modeling to capture complete facial geometry and texture information, which can then be computationally extracted to produce high-precision 2D facial images suitable for biometric recognition. This 3D-to-2D pipeline ensures accurate facial feature extraction while maintaining operational simplicity through automated processing.

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

Solution Approach 2:

The patent replaces manual image capture and selection processes with automated computational methods. The 3D head model is processed through software algorithms that automatically generate idealized 2D facial images and extract facial features, substituting mechanical photo capture operations with digital processing to achieve higher precision.

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

Data Source

PatentEP4064221A1Method for personalizing an id document and method for identifying a person using biometric facial features and id document
Publication Date: 2022.09.28 BUNDESDRUCKEREI GMBH
  • EP4064221A1 patent drawingFigure 1
  • EP4064221A1 patent drawingFigure 2
  • EP4064221A1 patent drawingFigure 3

AI summary

The invention relates to a method for personalizing an ID document comprising the following steps: providing image data for a person (4) in a data processing device (6), wherein the image data includes images with at least one facial image of the person (4); providing head model data representing a three-dimensional head model for the person (4), as well as texture data associated with the three-dimensional head model, in the data processing device (6); determining facial data representing a two-dimensional facial image with facial features for biometric facial recognition of the person (4) from the head model data and the texture data in the data processing device (6).Comprehensive determination of a two-dimensional projected facial image for person (4) by projecting the three-dimensional head model and determining the two-dimensional facial image for person (4) from the two-dimensional projected facial image, wherein at least one image section of the two-dimensional projected facial image is modified taking into account an associated image section in the at least one facial image capture. Furthermore, the facial data containing the two-dimensional facial image are stored in a storage medium (2) of an ID document (1). An ID document and a method for identifying a person based on biometric facial features are also provided. (Fig. 1)