Identity Document Image Processing for Forgery Resistance
Find Innovative SolutionsGenerate 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
Engineering 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
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.
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.
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
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.
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.
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
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.
Data Source
Figure 1A~3C
Figure 4A~5C
Figure 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.