Appearance segmentation and documentation applied to cultural heritage surfaces

This paper describes the development and application of a novel supervised segmentation technique used for conservation documentation based on visible appearance changes Cultural Heritage (CH) metal surfaces. The is employing linear discriminant analysis model to classify Reflectance Transformation Imaging (RTI) reconstruction coefficients. Hemispherical Harmonics (HSH) coefficients each pixel are first calculated then normalized. normalization increases robustness invariance making it possible apply documenting different surfaces at time intervals. In this paper, we presented three case studies related corrosion assessment CH objects through detection monitoring degree silver tarnishing. For study, data set constructed, teaching algorithm recognize as distinct specified characteristic (such corrosion, etc.) by comparing pixel. segmented information visualized simplified colormap. results afterwards verified inspection from conservation-restoration experts. method can segment with in micro-geometry, but reaches its limitation minimal topography high specularity.