Processing large ceramic assemblages using laser profiling and automated shape matching

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F67985912%3A_____%2F20%3A00532966" target="_blank" >RIV/67985912:_____/20:00532966 - isvavai.cz</a>

Result on the web

<a href="https://submissions.e-a-a.org/eaa2020/repository/preview.php?Abstract=1658" target="_blank" >https://submissions.e-a-a.org/eaa2020/repository/preview.php?Abstract=1658</a>

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Processing large ceramic assemblages using laser profiling and automated shape matching

Original language description

Capturing, cataloging and analyzing shapes of ceramic vessels is amongst the most informative, and also most laborious parts of the archaeological process. Advances in digital image processing and machine learning allow us to introduce an increasing degree of automation into pottery processing, simplifying the task of finding similarities and dissimilarities between individual vessels within assemblages counting often tens of thousands of fragments. To achieve good results, we need firstly digitization methods which allow us to capture the shapes of ideally all diagnostic fragments in a relatively short time frame. Secondly, we need methods which translate part of the expertise of a trained ceramicist into computer algorithms and pre-sort these large data sets into manageable typological groups. The Laser Aided Profiler (www.laseraidedprofiler.com), presented in this paper, allows for high-quality drawing documentation of dozens of ceramic fragments per day. It is a portable system, applicable e.g. at excavations abroad where long-term storage or transport of the finds to the home institution is not possible. Within a few minutes, it is possible to make a digital vector drawing that can be used to create a classic catalog and further use the acquired profile for morphometric analysis. We also present a new methodology of calculating morphometric (dis)similarity based on rim fragments, which was so far applied on 1200+ specimens of Bronze Age pottery. The resulting dissimilarity matrix served as input for Hierarchical Cluster Analysis, which was then employed in the typological classification of the material.

Czech name

—

Czech description

—

Type

O - Miscellaneous

CEP classification

—

OECD FORD branch

60102 - Archaeology

Project

<a href="/en/project/EF16_019%2F0000728" target="_blank" >EF16_019/0000728: Ultra-trace isotope research in social and environmental studies using accelerator mass spectrometry</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2020

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů