Assessment of the aging of ancient Chinese painting paper based on hyperspectral imaging technology

This study utilizes hyperspectral cube data collected from artificially aged Xuan paper samples, in combination with viscosity parameters that reflect the degree of paper aging, to establish a greedy partial least squares regression (PLSR) model for evaluating paper aging. The model was applied to assess the extent of aging of paper samples from Chinese calligraphy and paintings in the Nanjing Museum. The greedy PLSR model developed using artificially aged Xuan paper samples can effectively estimate the viscosity of Xuan paper used in Chinese calligraphy and paintings (with an error <10%), providing a highly reliable method for evaluating the aging of Xuan paper. However, the model shows larger estimation errors for other types of paper such as bamboo and machinemade paper, primarily due to differences in fiber structure and composition. Future research should expand the sample types and optimize spectral feature extraction and model adaptability to enhance the generalization capability for aging assessment of different paper materials.