Wedge Detection for Predictive Graphical Annotation of Cuneiform Tablets in 3D
No Thumbnail Available
Date
2024
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
As Digital Ancient Near Studies (DANES) and Ancient Language Processing (ALP) require larger volumes of annotated documents, we focus on automatically computed graphical annotations for cuneiform tablets, which are among the oldest documents spanning at least three millennia of human history. Cuneiform consists of wedges imprinted on clay tablets, which are best captured by high-resolution 3D acquisition. Existing Open Access 3D models are therefore our data base and have been postprocessed by the GigaMesh software framework, which ensures clean meshes and provides MSII filter responses. This refers to the Gaussian curvature, which is a function value for each vertex of the mesh used for watersheds along the 2D manifold. Watershed labels are determined by non-minimum suppression using a merge parameter. The label contours, i.e., the polygonal lines enclosing the wedges, are smoothed using the Savitzky-Golay filter with the goal of projecting a low-poly line in 2D that can be quickly corrected by the image-based annotation tool known as Cuneur, which allows experts to annotate wedges, signs and semantics. Once the wedges have been determined, wedge types can be automatically proposed using the orientation of the wedges, which is the first step towards a paleographic representation of cuneiform signs, known as PaleoCode.
Description
CCS Concepts: Computing methodologies → Computer graphics; Applied computing → Archaeology
@inproceedings{10.2312:gch.20241260,
booktitle = {Eurographics Workshop on Graphics and Cultural Heritage},
editor = {Corsini, Massimiliano and Ferdani, Daniele and Kuijper, Arjan and Kutlu, Hasan},
title = {{Wedge Detection for Predictive Graphical Annotation of Cuneiform Tablets in 3D}},
author = {Bullenkamp, Jan Philipp and Mara, Hubert},
year = {2024},
publisher = {The Eurographics Association},
ISSN = {2312-6124},
ISBN = {978-3-03868-248-6},
DOI = {10.2312/gch.20241260}
}