Chip Card Security via Multi-Resolution Image Feature Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current digital watermarking techniques for securing chip cards are fragile and fail to withstand damage such as print/scan distortions, geometric changes, and wear and tear, requiring database access for authentication, which complicates data security and storage.
Innovation Solution
A process that generates and matches image feature vectors from an image file and a scanned image signal using an insertion algorithm and read-back algorithm, ensuring a secure link between the microchip and printed image without adding physical or software elements, using a secret key for authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If digital watermarking techniques are used to secure chip cards, then the card security is improved, but the robustness of the security mark deteriorates due to print/scan distortions, geometric changes, and wear and tear
Solution Approach 1:
The patent divides the printed image into multiple sub-images at different resolution levels (coarse, medium, fine) and extracts feature vectors from each level. This segmentation allows the system to tolerate distortions at any single level by using feature vectors from other levels, thereby maintaining robustness against print/scan distortions, geometric changes, and wear and tear.
Solution Approach 2:
The patent introduces a multi-resolution dimension by processing the printed image at different resolution levels (coarse, medium, fine). This dimensional approach allows the system to extract feature vectors from multiple resolutions and use them for authentication, making the security mark robust against distortions that affect any single resolution level.
2Measurement precision
If database access is used for authentication, then the verification process is improved, but the data security and storage complexity deteriorates
Solution Approach 1:
The patent extracts and stores only the essential feature vectors (derived from image features at multiple resolutions) in the microchip rather than storing entire images or relying on external databases. This extraction approach maintains authentication accuracy while eliminating the need for complex database access and reducing storage requirements.
Solution Approach 2:
The patent creates a self-contained copy of the authentication data within the microchip itself, including feature vectors extracted from the printed image at multiple resolutions. This internal copying eliminates the need for external database access, simplifying the system while maintaining secure and accurate authentication.
3Reliability
If additional physical or software elements are added to the card, then the security measures are improved, but the card complexity and appearance alteration deteriorates
Solution Approach 1:
The patent makes the existing printed image serve multiple functions: it acts as both the visual identifier and the source of security data through extracted feature vectors at multiple resolutions. This multi-functionality eliminates the need for separate physical or software security elements, maintaining card simplicity while enhancing security.
Solution Approach 2:
The printed image itself provides the security functionality by having its feature vectors extracted and stored in the microchip. The system uses the image's own characteristics (at multiple resolutions) for authentication, eliminating the need for additional physical or software elements and keeping the card design simple.
Data Source
AI summary
The invention relates to a process to make secure a personal portable object comprising a body of the personal portable object, a microchip, a printed image and a device to enable said personal portable object to communicate with an entity external to the device. The process includes using an image file and an insertion algorithm to generate an image feature vector Vsi(num), storing the image feature vector Vsi(num) in the microchip, using the printed image obtained by a scanning device to generate an image feature vector Vsi(dig), and using a read-back algorithm to match the image feature vector Vsi(num) and the image feature vector Vsi(dig).