Degraded Artefact Image Data Set

The data set consists of 334 images of 1920x1080 pixels extracted from a video file of an artefact from the Restoration and Conservation Laboratory, Oltenia Museum, Craiova, Romania. The ones where the broken part is visible have associatated masks to cover the missing parts. The images can be used for image inpainting, as well as for 3D modelling. The video file included resulted from an approach that is presented in the article below, which is submitted for publication in Journal of Cultural Heritage.Ruxandra Stoean, Nebojsa Bacanin, Catalin Stoean, Leonard Ionescu, Bridging the Past and Present: AI-Driven 3D Restoration of Degraded Artefacts for Museum Digital Display, submitted to Journal of Cultural Heritage.If you publish material based on the current database, please refer the paper above within the References.The study took part within the project Object PErception and Reconstruction with deep neural Architectures (OPERA) https://sites.google.com/view/pce-opera/.

本数据集包含334张分辨率为1920×1080像素的图像，均提取自罗马尼亚克拉约瓦奥尔泰尼亚博物馆（Oltenia Museum）修复与保护实验室（Restoration and Conservation Laboratory）所拍摄的一件文物的视频素材。其中清晰呈现文物破损部位的图像，附带了用于标注缺失区域的掩码（mask）。该数据集可应用于图像修复（image inpainting）与三维建模（3D modelling）任务。
本数据集附带的视频素材源自下述已投稿至《文化遗产期刊》（Journal of Cultural Heritage）的文章中所提出的研究方法：
鲁克桑德拉·斯托安（Ruxandra Stoean）、内博伊沙·巴察宁（Nebojsa Bacanin）、卡塔林·斯托安（Catalin Stoean）、伦纳德·约内斯库（Leonard Ionescu），论文标题为《跨越古今：面向博物馆数字展示的、基于人工智能的受损文物三维修复》（Bridging the Past and Present: AI-Driven 3D Restoration of Degraded Artefacts for Museum Digital Display），该文已投稿至《文化遗产期刊》（Journal of Cultural Heritage）。
若基于本数据集开展研究并发表成果，请在参考文献中引用上述论文。
本研究依托深度神经网络目标感知与重建（Object PErception and Reconstruction with deep neural Architectures, OPERA）项目实施，项目官方网址为https://sites.google.com/view/pce-opera/。

搜集汇总

背景与挑战

背景概述

该数据集包含334张高分辨率（1920x1080像素）的退化文物图像，提取自一个博物馆文物视频文件，部分图像附带掩码以标识破损区域，专门用于图像修复和3D建模任务。数据集支持文化遗产保护领域的研究，尤其适用于深度学习和数字遗产应用，基于CC BY 4.0许可证开放共享。