Desarrollo de un método de clasificación de pastas cerámicas basado en el análisis de imágenes de radiografía digital

X-rays have motivated the application of imaging techniques in Heritage Sciences, providing a tool that produces global information about the object. The radiography analysis generates information with physical meaning relating the material properties with the photon attenuation. In this work, the characterization of a digital radiography system is presented, and the basis of an analysis method for ceramics sherds is described, based on the study of histograms, which represent the distribution of pixel intensities that made the image. Thus, this study aims to provide a tool, based on a mathematical analysis, encompassing a Lorentzian fit to each image histogram and a cluster analysis, that leads to establish a contrast between ceramic pastes. By applying this method, we could distinguish six paste groups from a set of 275 ceramic objects (archaeological and contemporary) from the Maya region.