Methodology for processing backscattered electron images. Application to Aguada archaeological paints.

Prior research has shown that scanning electron microscopy is widely used for material characterization, especially when X-ray analysis is not feasible. It is particularly valuable in analyzing thin layers, such as paint on ceramic artifacts, where electron interaction volumes may exceed the layer thickness. However, similar compositions between the paint and the ceramic substrate can obscure meaningful data. Established methods struggle with detecting subtle differences in mean atomic number, which are crucial for identifying chemical variations. This limitation hinders accurate analysis of archaeological materials. No prior work had resolved how to enhance these subtle contrasts effectively. That uncertainty drove the development of new image-processing techniques. Existing tools lack the ability to reveal small contrasts without significant detail loss. This gap motivated the creation of a specialized software to address the challenge. The study builds on prior knowledge of BSE imaging and mineralogical analysis of archaeological pigments.

The aim of this work is to develop and test a new image-processing methodology for backscattered electron (BSE) images. The goal is to enhance small mean atomic number contrasts that are typically imperceptible to the human eye. The specific problem is the difficulty in distinguishing between ceramic substrates and thin paint layers with similar compositions. The motivation arises from the need to complement traditional X-ray techniques in archaeological material analysis. The study focuses on black and white pigments from ceramic sherds of the Ambato style. The methodology is designed to preserve detail while enhancing chemical contrast. The approach is intended for routine use in analyzing samples with subtle differences. The software was tested on archaeological materials from the Aguada culture in Argentina.

The study involves the development of a custom image-processing software. The program implements a new methodology for BSE image treatment. The process begins with a smoothing filter to reduce noise in the images. After smoothing, a contrast enhancement routine is applied to highlight grey-level differences. An edge enhancement filter is then used to recover lost details. The software was tested on BSE images of ceramic sherds with black and white pigments. X-ray diffraction diagrams were collected for mineralogical validation. The Rietveld method was used with DIFFRACplus Topas software for analysis. The methodology combines spatial filtering and contrast enhancement techniques.

The software successfully revealed chemical contrasts between the ceramic body and pigments. BSE images initially showed no visible differences between the sherd and pigments. After smoothing, contrast enhancement revealed subtle grey-level variations. Edge enhancement filters recovered lost details without significant distortion. X-ray diffraction confirmed mineralogical differences between the pigments and the ceramic body. The results align with the BSE image analysis findings. The methodology proved effective in preserving detail while enhancing contrast. The software is suitable for routine analysis of samples with slight chemical differences.

The authors propose that the developed software is a suitable tool for analyzing archaeological materials. The methodology successfully enhances small mean atomic number contrasts in BSE images. The results suggest that the approach can be used in routine analysis of samples with subtle differences. The study confirms that X-ray diffraction validation supports the BSE image findings. The program is effective in preserving detail while enhancing chemical contrast. The authors suggest that the methodology complements traditional X-ray techniques. The findings indicate that the software is useful for analyzing thin paint layers on ceramics. The study highlights the potential of the software for broader applications in material characterization.