Digital image analysis on cathodoluminescence microscopy images for ancient ceramic classification: methods, applications, and perspectives

Abstract

This paper aims to investigate the application of cathodoluminescence (CL) imagery (with optical microscopy) and CL spectroscopy (with Scanning Electron Microscope) in ancient ceramics studies, for a proper classification of petro-fabrics and mineral inclusions based on their CL response. Digital image analysis (DIA) routines are proposed for both qualitative and quantitative analysis, and outcomes are reviewed in the light of classical optical microscopy (OM) and scanning electron microscopy (SEM–EDS) approaches. Obtained results demonstrated the suitability of CL in discriminating ceramic groups, offering also complementary information connected with aspects as provenance or manufacturing processes. CL imagery and spectroscopy data are also used to discuss perspectives of the method, debating bias, limits, or suspicious inconsistencies, exploring how compositional features or sample’s nature imprint CL response, and finally providing the basis for future developments. Addendum at the back of the paper includes and extensively discusses some technical aspects regarding the applied DIA routines, developed in open-sources software environments and available as supplementary materials.

Ceramic Microscope Image Classification Based on Multi-Scale Fusion Bottleneck Structure and Chunking Attention Mechanism

“…Hogan et al [13] discovered the relationship between compression testing and microstructure changes by conducting uniaxial and biaxial compression experiments on ceramics and conducting stress analysis while observing changes in ceramic microscopic images. Aprile et al [14], [15] identified the composition of ceramics through microscopic image acquisition methods such as OM and conducted modal analysis. This method of detection can avoid complex component extraction processes.…”

Inhomogeneous Beam Brightness Intensity Converter (IBBIC) as support software for cathodoluminescence microscopy studies and images pre-processing

“…The developed software has been tested on 412 images from scientific articles (Del Sole et al, 2020;Emami et al, 2021;Odelli et al, 2022). The pictures were acquired using CL microscopes with variable instrumental setting (100-300 µA, 7.5-15 kV, 15-72 s exposure time), and different typologies of surface treatment (polished and unpolished), therefore providing a method assessment related to a wide range of case studies.…”

“…Although CL results assessment might be an extremely delicate practice, it is widely employed since it is considered an outstanding method to detect both structural and chemical information about solids. The following are a few examples of the CL application in various fields of study: the investigation of building materials such as natural stone, cement, mortar and concrete (materials science) (Götze, 2009); the distinction between authigenic quartz precipitated as cement in sandstones and detrital quartz grains (petrography) (Sippel, 1968); the discrimination of ceramic groups, concurrently with information related to provenance or manufacturing processes (cultural heritage) (Odelli et al, 2022); the identification of rare earth element (REE)-bearing minerals in ore (mining sector) (Imashuku & Wagatsuma, 2021); the classification of zircons by combining them with deeps (mineralogy) (Zheng et al, 2022); the examination of the disparity in CL of quaternary carbonate deposits (sedimentary geology) (Braithwaite, 2016); the identification of meteorite impact microstructures and their demarcation from tectonic deformations ones (planetary sciences) (Hamers & Drury, 2011); the analysis of raw pharmaceutical materials and formulated products (pharmaceutical industry) (Nichols, 2012). Notwithstanding these cases are only an exiguous part among the countless bibliographic papers and they highlight the huge impact that CL has on solid materials studies.…”

“…Due to instrumentation geometry and the intensity of the electron beam, higher in its centre and gradually less intense moving away from it, analysed samples are subjected to different intensity of the ray incident. As a consequence, theoretically identical parts of the same sample would have a different response in terms of CL depending on the intensity of irradiation (Odelli et al, 2022). Although data assessments related to these typologies of study are often qualitative and this aspect is not significant, for a more quantitative approach it might be an obstacle to the interpretation.…”

“…This paper work starts from Odelli et al (2022), but it focuses mainly on the operator user experience, providing the operator with a set of tools capable of simplifying the choice of optimal parameters for the correction algorithm, thanks to the immediate, clear and precise feedbacks they provide. Thus, the IBBIC software speeds up the whole correction process and can also be used by operators who lack a wide proficiency in image processing theory and tools.…”