Digital Rubbing through Physical Simulation: A Non-Invasive Computational Approach for Epigraph Preservation

Abstract. The preservation of ancient Chinese inscription relics, including oracle bones, steles, and precious metal artifacts, faces critical challenges due to irreversible degradation from environmental erosion and physical contact during documentation. While traditional rubbing techniques remain widely used for recording surface inscription information, their mechanical ink-transfer process inherently causes cumulative damage through repetitive hammering and pressure application. This paper presents a novel computational framework through physical process simulation for non-invasive digital rubbing generation. Our key innovation lies in mathematically modeling the complex physical interactions involved in manual rubbing processes, including: (1) a Xuan paper generation model based on water retention properties, (2) a paper deformation model based on local depth, (3) an ink diffusion model based on contact probability. Experiments on ancient Chinese steles demonstrate our method’s superior visual fidelity over existing digital methods, particularly in restoring ink textures resembling actual rubbings.