Data

Data

For the development of the platform the consortium relied on several use cases of archaeological collections: votive terracotta statues from Salamis (Cyprus) and figurines from Ayia Irini (Cyprus), pottery fragments from Naukratis (Egypt), Tell es-Safi (Israel) and Pyla-Kokkinokremos (Cyprus), and fresco fragments from Antiphonitis Church (Cyprus). In total the project worked with over 450 objects with the majority coming from the Salamis dataset.

The datasets were documented in 3D with their colour information, further processed in various steps. Also their relative semantic information in form of museum records and archaeological descriptions was collected and mapped into the CIDOC CRM scheme.

Collection

# Objects

#3D models

Sources

Salamis

250

20 (Ashmolean)

190 (British)

19 (Cyprus)

12 (Fitzwilliam)

Ashmolean,

British Museum,

Cyprus Museum,

Fitzwilliam

Naukratis

~8000

​73 (British)

British Museum and project partners 

Antiphonitis Frescos

72

Department of Antiquities, Cyprus 

Pyla-Kokkinokremos jar

~10

Archaeological Museum of Larnaka 

Vessels of Tell es-Safi

The dataset of the 3D models were stored in a repository while the semantic information are gather in a triplestore, and are available under the Creative Commons Share-Alike license. To access the files, credentials are necessary. Credential, Triplestore and Repository can be accessed for free by clicking on the below icons:

Obtain the credentials

Triplestore download

The Triplestore

The semantic information are stored in a triplestore. The latest state of the triplestore serving the RDF content, is contained in a zipped journal file. The journal file can be used by instantiating a blazegraph triplestore.

The Repository

The repository hosts the dataset of the 3D models, except the Ayia Irini dataset that is not yet available due to research right restrictions. The structure of the repository is schematized in the picture.

On the first level of the repository there is a division according to museum and dataset, where each folder contains all the artefacts from that museum named after their Inventory Number. Each object contains one or two folders depending how many scans were carried out on that particular object. Each scan folder contains then the following: Images, Intermediate models and Processed models. Images contains relative png files and the Intermediate folder the final 3D model of the 3D documentation pipeline without further processing (“object.ID_final.ply”). The Processed models folder contains four different processed versions of the 3D model and two subfolders: “parts” and “properties”. The 3D models are the following: “object.ID_clean.ply”, which is the final 3D model cleaned with the specifically designed software component GraviFix and without colour, and three lower resolution models “object.ID_1M.ply”, “object.ID_100K.ply” and “object.ID_50K.ply”. The “parts” folder contains the respective faceting files for each resolution of the 3D models (e.g. “object.ID_100K.fct” ). Finally, the “properties” folder has five more files about geometric properties in it: “object.ID_1M_SI.txt” (Shape Index), “object.ID_1M_curvM.txt” (Mean Curvature), “object.ID_50K_SDF30.txt” (Shape Diameter Function), “object.ID_50K_mbb.json” (Minimal Bounding Box) and “object.ID_LAB.txt” (L-channel).