Quantitative analysis of authenticity-related features in Su Shi’s Gongfu Tie

Chinese calligraphy is an important component of cultural heritage, and authenticity assessment supports heritage preservation and calligraphic studies. Existing quantitative approaches often rely on single indicators, limiting their ability to capture structural proportion, ink distribution, and stroke boundary complexity. We propose a computer vision-based multidimensional framework integrating aspect ratio, black-to-white pixel ratio, and fractal dimension. By jointly characterizing structural, tonal, and boundary-complexity features, the framework captures complementary morphological information and is more suitable than single-feature approaches for detecting contour-regularization artifacts introduced by tracing or reproduction. Taking disputed Gongfu Tie as a case study, we constructed a reference dataset from authenticated works of Su Shi and analyzed morphological features using statistical comparison, K -means clustering, and cluster validity evaluation. The character “Shi” falls within the statistical ranges of the reference dataset and exhibits consistent clustering characteristics. The framework provides a reproducible and interpretable computational perspective for authenticity-related analysis of historical calligraphy.