A High-Accuracy SSIM-based Scoring System for Coin Die Link Identification

TL;DR

This work addresses the bottleneck of die-link identification in ancient coins by introducing a publicly available labeled image dataset (329 coins) and a novel SSIM-based distance that leverages global image information. After aligning coin images with SIFT-based keypoint overlays, the distance computes a mean SSIM-derived dissimilarity, and is benchmarked against Procrustes, FSIM, and VGG baselines. Evaluations using ROC/PR curves and clustering show near-perfect die-link identification for most datasets, with the SSIM-based approach also offering significantly faster computation. The results enable practical online deployment for large coin collections and lay a foundation for further enhancements in automated numismatics tools.

Abstract

The analyses of ancient coins, and especially the identification of those struck with the same die, provides invaluable information for archaeologists and historians. Nowadays, these die links are identified manually, which makes the process laborious, if not impossible when big treasures are discovered as the number of comparisons is too large. This study introduces advances that promise to streamline and enhance archaeological coin analysis. Our contributions include: 1) First publicly accessible labeled dataset of coin pictures (329 images) for die link detection, facilitating method benchmarking; 2) Novel SSIM-based scoring method for rapid and accurate discrimination of coin pairs, outperforming current techniques used in this research field; 3) Evaluation of clustering techniques using our score, demonstrating near-perfect die link identification. We provide datasets, to foster future research and the development of even more powerful tools for archaeology, and more particularly for numismatics.

Figures (7)

• Figure 1: Coins struck with the same die: example of ground truth on a dataset of 17 coins (DS8, as defined in Table \ref{['tab:dsslc']}).
• Figure 2: The Juillac treasure during the archaeological dig.
• Figure 3: Coin example for each dataset
• Figure 4: Pre-processing steps from the original coin to the input of the SSIM-based method (from left to right: raw database image; grayscale; cropping; CLAHE; denoising).
• Figure 5: Steps of the computation of the SSIM-based distance, and comparison of two results (one linked, one unlinked). The greener the color (low distance), the more likely it is to be a die link. The related algorithm is detailed in Algorithm \ref{['algo:ssim']}.