Identity Document Image Processing for Forgery Resistance

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The challenge in identity document security is the complexity of falsifying portrait images, particularly with 'morphed' images that resemble the legitimate bearer, making it difficult to distinguish authentic from forged documents, even with automatic controls.

Innovation Solution

An image processing method that decomposes a source image into gray levels, applies error diffusion dithering to convert it into a raster image in black and white, and encodes information using a correspondence table to reliably prove the authenticity of a portrait, incorporating error correcting codes and digital signatures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If morphed images are used to create forged identity documents, then the visual resemblance to the legitimate bearer is improved, but the reliability of the document is worsened

Engineering Contradiction:
Improvevisual resemblanceVSAvoiddocument authenticity
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent divides the portrait image into multiple small blocks arranged in a grid pattern. Each block is further subdivided into smaller pixels that represent different gray levels. This segmentation allows the embedding of authentication data within the image structure itself, making it difficult to forge while maintaining visual quality.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing to different regions of the image by treating each block independently. The error diffusion dithering algorithm processes each block locally, preserving local visual characteristics while embedding authentication information. This local processing ensures that the authentication data is distributed throughout the image rather than concentrated in one area.

Inventive Principle:
Principle #3Local quality

2Reliability

If error diffusion dithering is applied to encode authentication information, then the reliability of the document is improved, but the device complexity is worsened

Engineering Contradiction:
Improveauthentication accuracyVSAvoidprocessing algorithm complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent pre-defines a correspondence table that maps between block patterns and authentication data values before the actual encoding process. This preliminary preparation simplifies the main encoding operation, as the system only needs to look up pre-computed patterns rather than performing complex calculations in real-time. The correspondence table stores the relationship between different block configurations and their corresponding authentication information.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent uses error diffusion dithering, which is a well-established image processing technique already widely used in printing and display technologies. By leveraging this existing, well-understood algorithm, the patent avoids the need to develop and implement a completely new complex processing method. The error diffusion process naturally distributes quantization errors across neighboring pixels, which inadvertently embeds authentication information in the image structure.

Inventive Principle:
Principle #26Copying

3Reliability

If the portrait image is decomposed into blocks with encoded information, then the security against forgery is improved, but the ease of operation is worsened

Engineering Contradiction:
Improveforgery resistanceVSAvoidverification process simplicity
Core Design Contradiction:
ReliabilityVSEase of operation

Solution Approach 1:

The patent embeds authentication data directly within the portrait image itself, rather than requiring separate authentication mechanisms. The image serves dual purposes: it is both the visual representation of the person and the carrier of authentication information. This self-contained approach simplifies verification, as the same image used for visual identification also provides the authentication data, eliminating the need for additional separate verification steps.

Inventive Principle:
Principle #25Self-service

Data Source

PatentEP3683729B1Image-processing method for identity document
Publication Date: 2022.09.07 SURYS
  • EP3683729B1 patent drawingFigure 1A~3C
  • EP3683729B1 patent drawingFigure 4A~5C
  • EP3683729B1 patent drawingFigure 6A~8A

AI summary

The invention relates to an image processing method for identity documents. It is essentially characterized in that it comprises steps consisting of: • Decomposing a grayscale source image into a set of rectangular blocks of pixels, each block comprising at least 4 pixels, • Applying an error diffusion dithering algorithm to the source image decomposed into blocks, then for each block: • Determining the average light intensity of the block, • Selecting from a set of predetermined cells, a cell that has an average light intensity corresponding to the light intensity of the block, - said set of predetermined cells comprising a plurality of individual cells, - each individual cell comprising a set of at least 4 pixels representing a unique pattern stored in memory, and • Replacing the block with the selected cell.