Encoding Hidden Images in Host Image Spatial Frequencies
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for encoding hidden images in secure documents, such as digital watermarking in binary halftoning, face challenges with precision and sensitivity to artefacts, limiting their robustness and practicality for counterfeiting prevention.
Innovation Solution
The method involves encoding hidden information in the spatial frequencies of high-frequency spatial features of a host image, which does not rely on optical density or luminosity measurements, allowing for a visible image to be incorporated without interfering with the hidden information, using a periodic line texture or stochastic line patterns, and employing specific mappings and decoding techniques to extract the hidden image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If digital watermarking is encoded in a digital image using binary halftoning, then the hidden information can be incorporated into the document, but the precision and robustness of the hidden information is reduced due to limitations of the printing method
Solution Approach 1:
The patent transitions from encoding hidden information in amplitude/optical density (traditional binary halftoning) to encoding in spatial frequency domain. This dimensional change allows the hidden message to be represented by the orientation and frequency of line patterns rather than by pixel intensity variations, making it resilient to printing artifacts that affect amplitude but preserve frequency characteristics.
Solution Approach 2:
The invention changes the encoding parameter from optical density/amplitude to spatial frequency characteristics (line spacing, orientation, frequency). By modulating these frequency parameters in high-frequency regions of the host image, the system achieves robust encoding that survives the printing process while maintaining the ability to recover the hidden message with high precision.
2Reliability
If the hidden information is encoded using traditional watermarking methods, then the document security is enhanced, but the visible quality of the image may be degraded or the hidden information may be detectable
Solution Approach 1:
The patent applies different characteristics to different regions of the image by encoding the hidden message specifically in high-frequency spatial regions where human visual sensitivity is reduced. The line patterns are embedded in textures and fine details that are imperceptible to the human eye, making the hidden information undetectable while maintaining overall image quality.
Solution Approach 2:
The invention converts the limitation of binary halftoning (coarse quantization) into an advantage by using the inherent high-frequency noise and texture patterns created by halftoning as the carrier for the hidden message. Rather than trying to overcome these artifacts, the method embeds the secret information within them, making the hidden data resilient to further processing and copying.
3Reliability
If the hidden information is encoded in the host image, then the document authenticity can be verified, but the complexity of the encoding and decoding process increases
Solution Approach 1:
The patent replaces complex iterative optimization algorithms and artificial intelligence-based encoding methods with a more straightforward frequency-domain modulation approach. By using spatial frequency analysis and line pattern synthesis, the system achieves robust hidden message embedding with simpler, more deterministic processing that is easier to implement and verify.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A method of encoding a hidden image in high frequency spatial frequencies of a line pattern of a host image. A set of host image spatial frequencies is generated based on a predefined mapping of a domain of a set of representative scalar values of the hidden image and a domain of the host image spatial frequencies. The line pattern of the host image is generated based on the set of host image spatial frequencies. The host image may be composed of tiles containing parallel line segments, with each tile encoding a corresponding one of the scalar values. The host image may be composed of a stochastic line pattern generated from a white noise image convolved with a space variable kernel based on the predefined domain mapping. The hidden image may be decoded algorithmically or optically in a single step.