CT reconstructions, analytical outputs and 3D shape data for Únětice pottery from eastern Bohemia

Submitted by Richard Thér on
Subtitle
HOR9
Is Part of Dataset Collection
CT reconstructions, analytical outputs and 3D shape data for Únětice pottery from eastern Bohemia
Resource Type
dataset
Creators and Contributors
Creator
Richard Thér ; University of Hradec Králové, Department of Archaeology richard.ther@uhk.cz
Publisher
ArchaeoVault
Description
Purpose and Context of Dataset Creation

This dataset was created as the source and documentation dataset for the study Pottery-forming practices in the Early Bronze Age Únětice culture of eastern Bohemia: CT-based analysis of technological variability at the microregional scale. The dataset documents ceramic vessels and vessel fragments from four Early Bronze Age Únětice culture sites in the Hradec Králové region of eastern Bohemia: Plotiště nad Labem A, Plotiště nad Labem B, Lochenice, and Horní Černůtky.

The purpose of the dataset is to provide primary and derived data supporting the analysis of pottery-forming practices, technological variability, and vessel-part-specific forming routines. The dataset was created to enable reproducibility of the CT-based structural evaluation, 3D pore-orientation analysis, 3D surface documentation, and profile-based morphometric analysis used in the associated study.

The dataset includes non-destructive CT reconstruction data, derived CT section images, segmented pore/fabric measurements, 3D geometry associated with segmented pore data, structured-light 3D surface scans, and outputs of vessel-profile morphometric analysis. It therefore documents both the internal structure and external shape of the analysed ceramic samples.

Methods

The dataset was created using a combination of non-destructive computed tomography, section-based structural evaluation, 3D segmentation of pore/fabric components, structured-light 3D scanning, and profile-based morphometric analysis.

CT imaging was used to document the internal structure of the ceramic samples. The reconstructed CT volumes were examined using radial sections and non-planar parallel sections following vessel-wall curvature. These sections were used to evaluate structural discontinuities, joint morphology, pore organisation, and fabric patterns relevant to the reconstruction of pottery-forming techniques.

Quantitative pore-orientation analysis was based on segmentation of 3D pore representations in selected vessel parts. For each analysed vessel part, a selected surface area of approximately 400–600 mm² was used for pore extraction and orientation analysis. Segmented objects were measured using variables such as volume, centroid coordinates, principal-axis direction, roundness, elongation, and bounding-box dimensions. Objects with elongation greater than 3 were used for the orientation statistics.

Orientation analysis was performed using an R-based workflow. The workflow processed CSV files containing segmented pore measurements, fitted parabolic approximations of vessel-wall curvature, projected pore principal axes into wall-related analytical planes, and calculated axial orientation statistics. The main outputs include mean axial direction, deviation from the expected horizontal or wall-parallel reference direction, circular standard deviation, Rayleigh test statistic, Rayleigh test p-value, and classification of orientation patterns in the parallel and radial planes.

The external shape of ceramic samples was documented using structured-light 3D scanning. Profile-based morphometric analysis was performed on vessel profiles converted into vector representations. The workflow extracted key landmarks, including maximum body diameter, rim diameter, and maximum neck constriction, and calculated vessel shape descriptors and volume estimates under the assumption of axial symmetry.

Series Information

The dataset is not a time series or monitoring series. It is a sample-based analytical dataset composed of repeated documentation and analytical outputs generated for individual ceramic samples and, where applicable, for individual vessel parts.

The repeating structure of the dataset follows the analytical division of each vessel or vessel fragment into morphological zones, especially lower body/base, mid-body, and neck. For each sample and vessel part, related files may include segmented pore measurements, orientation-analysis visualisations, and 3D geometry outputs. This structure reflects the analytical design of the associated study, in which vessel parts are treated as meaningful units because they may have been formed, transformed, or finished differently during manufacture.

The dataset is linked to the supplementary materials of the associated publication, especially the documentation of analysed archaeological samples from Plotiště nad Labem A, Plotiště nad Labem B, Lochenice, Horní Černůtky, the experimental reference dataset, and the database of analysed archaeological samples.

Dataset Contents and Structure

The dataset contains primary CT reconstruction data, derived section images, segmented pore measurements, 3D surface models, 3D geometry outputs related to segmented pores, and morphometric visualisations.

The main file types and their functions are as follows:

  • sample.raw: binary 3D CT reconstruction volume. This is the primary volumetric CT data file.
  • sample.raw.txt: metadata file required for correct interpretation of the raw CT volume. It contains information such as voxel dimensions, data type, data-type size, voxel size or spacing, and, where present, spatial orientation information.
  • sample.obj: 3D surface model of the ceramic sample generated by structured-light 3D scanning. It documents the external surface geometry of the sample.
  • sample_base.csv, sample_body.csv, sample_neck.csv: semicolon-separated CSV files containing measurements of segmented pores or fabric components from the lower body/base, mid-body, and neck zones of the sample.
  • sample_base.pdf, sample_body.pdf, sample_neck.pdf: graphical outputs generated by the R orientation-analysis workflow from the corresponding CSV files. They visualise pore-vector orientations and interpolated axial orientation fields.
  • sample_base.ply, sample_body.ply, sample_neck.ply: 3D geometry files associated with segmented pore/fabric data for individual vessel parts.
  • sample_par.jpg: example non-planar parallel CT section following the curvature of the vessel wall.
  • sample_rad.jpg: example radial CT section through the vessel wall.
  • sample_XY3D.png: profile-derived XY projection or 3D morphometric visualisation generated from the vessel-profile analysis workflow.

The CSV files contain 22 columns describing each segmented pore or pore-like object. These include voxel count, object volume, surface area, compactness/sphericity index, centroid coordinates, principal-axis vector components, roundness, elongation, Willmore energy, distances from reference planes, and bounding-box voxel indices.

The structure of the dataset is designed to preserve the relationship between primary data and derived outputs. The raw CT volume is the primary internal-structure record; section images are selected visual exports from the CT volume; CSV files contain quantitative segmented-pore measurements; PDF files visualise the R-based orientation analysis; PLY files store 3D geometry associated with the segmented pore/fabric data; OBJ files document the external surface geometry; and PNG files document outputs of profile-based morphometric analysis.

Technical Information

The CT data were acquired using an Explorer X test 150/100 system. The instrument is a customised, non-serial configuration developed to meet customer-specific requirements. The system includes an X-ray generator with a 0.05 mm focal spot. Scans were performed at 145 kV and 0.21 mA. Projections were recorded using a flat-panel detector with an active area of 301 × 250 mm, 3008 × 2496 pixels, and a pixel size of 0.10 mm. The achievable spatial resolution was as low as 25 µm for a field of view of approximately 60 × 70 mm. The maximum object size was approximately 23 × 19 cm, yielding a voxel size of approximately 80 µm. No beam filtration was applied.

The raw CT files are binary volumes and require the accompanying .raw.txt metadata files for correct interpretation. These metadata files should be consulted before opening the raw data, especially to determine data type, voxel matrix dimensions, voxel size, byte order, and any spatial orientation information.

The segmentation and orientation analysis of pore/fabric components were carried out in LametomArk and through a custom R-based workflow. The segmentation model classified CT volume voxels into background, ceramic matrix, and pore classes. For each segmented pore, object measurements including centroid coordinates and covariance-based principal-axis direction were calculated. Orientation statistics and plots were then generated in R.

The R workflow used parabolic approximations of vessel-wall curvature to express pore orientations relative to wall-related reference planes. Axial orientation statistics were calculated by doubling axial angles, computing circular statistics in directional space, and transforming the results back to axial orientation. The workflow used R packages including tidyverse for data handling and plotting, and base R/statistical functions for circular calculations.

The 3D surface models were produced by structured-light scanning using an Artec Space Spider device, with accuracy down to 0.05 mm and 3D resolution of 0.1 mm. The scans were processed in Artec Studio 18 software.

The OBJ and PLY files can be inspected in software such as MeshLab, CloudCompare, or Blender. The raw CT data can be opened in Fiji/ImageJ using the parameters from the corresponding .raw.txt metadata file. If exact spatial orientation metadata are required, conversion to a format such as NRRD/NHDR and inspection in software such as 3D Slicer is recommended.

Research Questions

The dataset was created to address the following research questions:

  1. Which pottery-forming techniques and forming methods were used in the Early Bronze Age Únětice ceramic assemblages from eastern Bohemia?
  2. Does the analysed pottery show technological variability at the microregional scale, and can this variability be related to chronological phase, site, vessel type, or vessel part?
  3. Do individual vessel parts, such as necks, mid-bodies, lower bodies, and bases, preserve different structural patterns that reflect different forming routines?
  4. Can CT-based section analysis and quantitative 3D pore-orientation analysis improve the reconstruction of hand-building practices in archaeological ceramics?
Appropriate Reuse

The dataset is appropriate for reuse in research on archaeological ceramic technology, pottery-forming practices, CT-based analysis of ceramic fabrics, 3D pore-orientation analysis, and morphometric analysis of vessel shape.

The data may be reused for:

  • verification and reproducibility of the analyses presented in the associated publication;
  • comparative studies of pottery-forming techniques in Early Bronze Age or other hand-built ceramic assemblages;
  • methodological development of CT-based ceramic analysis;
  • testing or refining workflows for segmentation and orientation analysis of pores or fabric components;
  • comparison of internal structural features with vessel morphology and vessel-part organisation;
  • teaching and demonstration of non-destructive imaging methods in archaeological ceramic studies;
  • visualisation of internal and external ceramic structure using CT volumes and 3D models;
  • integration with other archaeological, archaeometric, or experimental reference datasets, provided that differences in acquisition parameters, segmentation procedures, and sampling design are taken into account.
Dataset limitations and inappropriate reuse

The dataset has several limitations that should be considered in any reuse.

The analysed assemblage is selective and reflects the sampling strategy of the associated study. It was designed to capture morphological variability, principal vessel parts, and selected chronological and contextual contrasts in the Hradec Králové region. It should not be treated as a complete inventory of all Únětice pottery from the region.

Not all samples preserve all vessel parts. Some data are available only for specific zones, such as the lower body/base, mid-body, or neck. Comparisons between vessel parts must therefore take differences in preservation, sampling, and analytical coverage into account.

The JPEG section images are illustrative exports only. They do not constitute exhaustive documentation of all diagnostic structural features. Technological interpretation should not be based on a single section image alone, but should take into account the full CT volume, multiple section positions, and the derived quantitative outputs.

The CSV, PDF, and PLY files are derived analytical outputs based on segmentation of pore/fabric components. They depend on the segmentation workflow, filtering criteria, and analytical assumptions used in the study. They should not be interpreted independently of the CT volumes and methodological documentation.

The dataset is not appropriate for:

  • provenance analysis of ceramic raw materials;
  • chemical, mineralogical, or petrographic characterisation unless combined with additional analytical data;
  • direct dating of archaeological contexts;
  • population-level statistical generalisation beyond the sampled assemblage without additional data;
  • functional interpretation of vessels without considering archaeological context and morphology;
  • high-precision metrological use beyond the documented resolution, accuracy, and processing conditions;
  • reconstruction of complete vessel forms where only fragments or partial profiles are documented;
  • automated classification of pottery-forming techniques without expert evaluation of CT features and methodological constraints.

The dataset should also not be used as a standalone typological or chronological database of Únětice pottery. Its primary purpose is technological and morphometric documentation related to the CT-based analysis of pottery-forming practices.

Time Reference
Time Period
Únětice Culture
Date Collected
2023-01 to 2025-12
Date Created
2026
Publication year
2026
Subject

Keywords

Assessment Type
material analysis
Object Type
potsherd
Object Specification
ceramic
Feature Specification
ceramic
Feature Type
sunken feature
Funding
Funder
Name of organisation
Grantová agentura ČR
Link to funding program
https://gacr.cz/zakladni-informace/
Award title
Pottery as a Witness to Cultural Change? The Early Bronze Age Settlement Agglomeration in Plotiště n. Labem in the Light of Multidisciplinary Research
Local identifier
23-07619S
Instrument

Explorer X test 150/100 system

Instrument ID
Not specified in the available documentation.
Instrument type
Computed tomography / X-ray imaging system
Manufacturer
Testima
Model
Explorer X test 150/100
Instrument owner
Universtity of Hradec Králové

Artec Space Spider

Instrument ID
Not specified in the available documentation.
Instrument type
Structured-light 3D scanner
Manufacturer
Artec 3D
Model
Space Spider
Instrument owner
Universtity of Hradec Králové
Location
Refers to the location
Plotiště nad Labem
Location ID
AMČR achaeological site
C-202310217
Refers to the location
Lochenice
Location ID
AMČR achaeological site
C-202008729
Refers to the location
Horní Černůtky
Location ID
AMČR achaeological site
C-202501334
Dataset Language
English
License
Creative Commons Attribution 4.0 International
Displaying dataset files 1 - 8 of 8
Name File size
HOR9.obj
HOR9.raw
HOR9.raw.txt 576 bytes
HOR9_body.csv 427.85 KB
HOR9_body.pdf 107.14 KB
HOR9_body.ply
HOR9_par.jpg 415.02 KB
HOR9_rad.jpg 146.71 KB