Empirical Mass-Retention Modeling, Image-Based Surface Texture Analysis and Color Evaluation in Vegetable Retanning of Wet-Blue Leather

Vegetable retanning modifies the surface, mass-retention and initial color-difference characteristics of wet-blue leather, but dosage selection is still commonly based on practical experience rather than quantitative process descriptors. This study developed a low-cost empirical screening framework for wet-blue bovine leather retanned with mimosa, quebracho and their 50:50 binary mixture at pH 4.0 and pH 5.5. Forty initial tannin dosage levels between 0.5 and 20 g/L were evaluated. Because equilibrium bath concentration was not measured, tannin uptake was expressed only as gravimetric dry weight gain, and fitted curves were interpreted as initial-dosage-based empirical responses rather than thermodynamic adsorption isotherms. Initial color response was evaluated using CIEDE2000 color difference, while digital surface images were analyzed using gray-level co-occurrence matrix descriptors and Surface Roughness Entropy. Quebracho at pH 4.0 displayed a bounded mass-retention trend associated with increased two-dimensional image regularity and reduced entropy. Mimosa at pH 5.5 displayed sustained mass accumulation accompanied by localized image-texture heterogeneity. Across the tested range, mimosa produced lower initial color-difference values than the other tannin systems, while the binary mixture showed intermediate behavior. The combined gravimetric, colorimetric and image-based approach provides a practical non-destructive screening strategy for comparing vegetable retanning conditions. The findings should be interpreted as comparative process trends and should be validated in future work by bath exhaustion monitoring, shrinkage temperature, standardized physical tests and spatially calibrated imaging.