Developing an automatic process for the segmentation and dimensional characterization of high-semantic level features from a ceramic find is an essential prerequisite for obtaining faster, reproducible, and more accurate measurements than the manual approach. These measurements are essential for analyzing, interpreting, and classifying the archaeological pottery, comparing and analyzing similarities, identifying the presence of standard attributes in the ceramics recovered from a specific archaeological site, or studying ancient manufacturing technologies. This paper proposes a new methodology for the recognition and dimensional measurement of a specific class of geometric features starting from high-density tessellated models acquired by 3D scanners, the Constant-Radius Sweeping Features (CRSFs). The recognition process is performed based on a fuzzy algorithm, which aggregates similar adjacent nodes, according to values of appropriate membership functions, into a single geometric feature. CRSFs are frequently seen in ancient artifacts as convex traces on the ceramic surface, such as plastic and molded reliefs, or concave features, such as engravings, graffiti, working signs, and impressions/stampings. Although they are frequently characterized, from a geometric point of view, by free-form surfaces, CRSFs may also be axially symmetrical geometry: this occurs quite often in archaeological pottery in correspondence with rims, bases, or external walls. In the proposed experimentation, the new methodology is applied to three fragments belonging to the same ceramic vessel and sharing a part of its rim. The results show that the algorithmic implementation of rules for CRSF recognition and measurement enables the automation of the entire process, from feature segmentation to the evaluation of the relevant characteristic dimensions, with the benefit of obtaining more robust and precise measurements than those performed manually. Furthermore, in some circumstances, the methodology proposed here allows for assessing dimensional attributes that would otherwise be impossible to evaluate by conventional methods: this is the case of CRSF not attributable to analytical geometric types, as frequently occurs in archaeological ceramics in the form of decorations, grooves, and processing marks.

Development of an automatic methodology for the recognition and dimensional characterization of constant-radius sweeping features from ancient ceramic artefacts

Di Angelo L.;Di Stefano P.;Guardiani E.;
2023-01-01

Abstract

Developing an automatic process for the segmentation and dimensional characterization of high-semantic level features from a ceramic find is an essential prerequisite for obtaining faster, reproducible, and more accurate measurements than the manual approach. These measurements are essential for analyzing, interpreting, and classifying the archaeological pottery, comparing and analyzing similarities, identifying the presence of standard attributes in the ceramics recovered from a specific archaeological site, or studying ancient manufacturing technologies. This paper proposes a new methodology for the recognition and dimensional measurement of a specific class of geometric features starting from high-density tessellated models acquired by 3D scanners, the Constant-Radius Sweeping Features (CRSFs). The recognition process is performed based on a fuzzy algorithm, which aggregates similar adjacent nodes, according to values of appropriate membership functions, into a single geometric feature. CRSFs are frequently seen in ancient artifacts as convex traces on the ceramic surface, such as plastic and molded reliefs, or concave features, such as engravings, graffiti, working signs, and impressions/stampings. Although they are frequently characterized, from a geometric point of view, by free-form surfaces, CRSFs may also be axially symmetrical geometry: this occurs quite often in archaeological pottery in correspondence with rims, bases, or external walls. In the proposed experimentation, the new methodology is applied to three fragments belonging to the same ceramic vessel and sharing a part of its rim. The results show that the algorithmic implementation of rules for CRSF recognition and measurement enables the automation of the entire process, from feature segmentation to the evaluation of the relevant characteristic dimensions, with the benefit of obtaining more robust and precise measurements than those performed manually. Furthermore, in some circumstances, the methodology proposed here allows for assessing dimensional attributes that would otherwise be impossible to evaluate by conventional methods: this is the case of CRSF not attributable to analytical geometric types, as frequently occurs in archaeological ceramics in the form of decorations, grooves, and processing marks.
File in questo prodotto:
Non ci sono file associati a questo prodotto.
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11697/203759
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? 0
social impact