Bayesian Muography Incorporating Prior Knowledge: A High-Resolution Imaging Algorithm for Structurally Complex Cultural Artifacts

Cultural artifacts are vital for heritage preservation but vulnerable to environmental damage that creates internal structural defects not visible on the surface. The drainage dragon heads of the Forbidden City exemplify this problem, suffering from structural detachment due to long-term erosion, which threatens both heritage preservation and public safety. Conventional non-destructive testing methods are inadequate in this context. Contact-based ultrasound risks surface damage, while X-ray devices are unsuitable for public heritage sites and both lack sufficient penetration capability. Muography offers a promising non-invasive solution using cosmic ray muons to image internal structures without artificial radiation. However, conventional muography algorithms lack the resolution needed for fracture detection in drainage structures. We propose a highprecision scattering-based muography algorithm using Bayesian inference with prior knowledge. This approach incorporates the object’s geometric outline and constructs a likelihood function combining theoretical scattering models with muon momentum. Using PyMC for probabilistic sampling, we reconstruct slice density to reveal internal structures. Both simulation and experimental results demonstrate significant improvements in imaging complex fractures in drainage structures. This method provides a practical solution for cultural heritage preservation and can extend to structural assessment in civil engineering applications.