Search

This paper presents an algorithm for automated detection and classification of multi-cell phytoliths, one of the major components of many archaeological and paleoenvironmental deposits. This identification, based on phytolith wave pattern, is made using a pretrained VGG19 deep learning model. This approach has been tested in three key phytolith genera for the study of agricultural origins in Near East archaeology: Avena , Hordeum and Triticum. Also, this classification has been validated at species-level using Triticum boeoticum and dicoccoides images. Due to the diversity of microscopes, cameras and chemical treatments that can influence images of phytolith slides, three types of data augmentation techniques have been implemented: rotation of the images at 45-degree angles, random colour and brightness jittering, and random blur/sharpen. The implemented workflow has resulted in an overall accuracy of 93.68% for phytolith genera, improving previous attempts. The algorithm has also demonstrated its potential to automatize the classification of phytoliths species with an overall accuracy of 100%. The open code and platforms employed to develop the algorithm assure the method's accessibility, reproducibility and reusability. • Deep learning automates identification and classification of multi-cell phytoliths. • Can be applied to digitalised slides and integrated into digital microscopes. • Can differentiate between phytolith genera but also species. • Presents an overall accuracy of 93% and analyses the whole slide. • Provides complementary metrics with new relevant information on phytoliths. [ABSTRACT FROM AUTHOR]

Seals and sealings were widely used in Neopalatial Crete, primarily in community-wide administrative systems to manage the movement and storage of goods, but also to secure or control channels of intra-site communication using parchment documents. The main principle of this monitoring process was the use of the seal impression as an 'indexical' sign, that is, as a means of identifying a seal-user who was acting in an authorising or securing role. Given the fact that seal-users could be acting on their own behalf or, alternatively, as representatives of an institution, there is uncertainty in the literature still over whether the seal impression stood for a single identifiable individual or not. In this paper, we present a new computational methodology for identifying and quantifying similarity within sets of multiple seal impressions made by the same seal. Using a combination of visualisations derived from computational analysis of 3-D scans, experimental work to make seal impressions, and scrutiny of the relevant sealing casts within the Corpus of Minoan and Mycenaean Seals, Heidelberg, we are able to identify groups of sealings impressed in significantly similar ways. We consider it probable that seal impressions with a high degree of similarity were made by the same individual or by two or more people trained to work in a very specific way. This computational methodology is transferable and would prove valuable for other collections of material with identifiable repeated visual elements, such as coins. The implications of our research and the nuance it adds to reconstructions of Neopalatial sealing practices are of broader relevance for sphragistic studies, and provide a point of comparison with the contemporary sealing material of the Ancient Near East. • 3D scanning and computational analysis identify significant differences in repeated impressions. • Visually similar impressions suggest training for careful maintenance of practice. • A seal was probably not always used by the same person in Neopalatial Crete. [ABSTRACT FROM AUTHOR]