Surface characterization of Asian lacquers using surface metrology and data science: introducing the roughness spectrum

Abstract A quantitative approach to the study of Asian lacquer surfaces combining non-contact, non-invasive, and non-destructive surface metrology and data science techniques is presented. The lacquers within this artistic tradition–laccol, thitsi, and urushi– combined with various additives possess quantifiable differences in the surface texture/topography that may be used to detect and identify a lacquer type from non-perturbing/contact measurements. This study examined laccol, thitsi, and urushi lacquer handmade test panels with various oils, pigments, and resins before and after aging. Confocal microscopy was used to acquire quantitative surface texture areal data from the test panels and a set of works of art and cultural heritage objects. Data science methods of feature engineering and convolutional neural networks (CNN) were applied to analyze the numerical surface texture data, assign lacquer specimens to the three lacquer types, and quantify the surface changes associated with aging.