Assessing the Viability of Synthetic Physical Copy Detection Patterns on Different Imaging Systems

TL;DR

The research demonstrates that synthetic CDP offer substantial improvements in authentication accuracy compared to one based on traditional digital templates t, and indicates that synthetic CDP can reliably differentiate between original and fake samples, making this approach a viable solution for real-world applications.

Abstract

This paper explores the potential of synthetic physical Copy Detection Patterns (CDP) to improve the robustness of anti-counterfeiting systems. By leveraging synthetic physical CDP, we aim at enhancing security and cost-effectiveness across various real-world applications. Our research demonstrates that synthetic CDP offer substantial improvements in authentication accuracy compared to one based on traditional digital templates. We conducted extensive tests using both a scanner and a diverse range of mobile phones, validating our approach through ROC analysis. The results indicate that synthetic CDP can reliably differentiate between original and fake samples, making this approach a viable solution for real-world applications, though requires an additional research to make this technology scalable across a variety of imaging devices.

Figures (6)

• Figure 1: The diagram demonstrating the real-world CDP lifecycle which includes the authentication of probe ${\bf y}$, the possible attack scenario (red dashed line) yielding fake ${\bf f}$, and the proposed defensive strategies (green dashed line) based on the reference ${\bf r} \in \{{\bf t}, { \mathbf{x}_e }, { \mathbf{\hat{x}} }\}$.
• Figure 2: The distributions of ${ f_{\mathrm{pcorr}} }({\bf y}, {\bf t})$ and ${ f_{\mathrm{ssim}} }({\bf y}, {\bf t})$ for both printers, where probes $\bf y$ were scanned with a flatbed scanner. $\bf f$ and $\bf x$ denote fake and original probes, respectively, and $\bf t$ stands for the digital template.
• Figure 3: The examples of the left-upper crop of CDP #166 printed on HPI55 and captured with a whole range of studied devices. Along with the digital template ${\bf t}$ and its printed form ${\bf x}$, the figure comprises the synthetic physical ${ \mathbf{\hat{x}} }$ that visually closely resembles real physical CDP ${\bf x}$.
• Figure 4: The scatter plots of both ${ f_{\mathrm{sim}} }$ for all given combinations of printer and imaging device.
• Figure 5: The distributions of ${ f_{\mathrm{pcorr}} }({\bf y}, {\bf r})$ for CDP captured with the worst-case imaging system, iPhone XS wide.