Image Segmentation in Loss Detection for Conservation and Restoration of Old Paintings

20203 citationsJournal Article

Authors

Roman Mykolaichuk · Taras Shevchenko National University of Kyiv

Vladyslav Lavrynovych · Taras Shevchenko National University of Kyiv

Antonina Mykolaichuk · National Academy of Fine Art and Architecture

Tetiana Tymchenko · National Academy of Fine Art and Architecture

Iryna Somyk-Ponomarenko · National Academy of Fine Art and Architecture

Abstract

Old paintings are usually characterized by presence of damaged areas. This article presents technologies and methods for detection of the loss on old paintings using image processing, image segmentation techniques, and convolutional neural network for detection of damaged parts on painting. Analysis of model prediction results allows us to estimate the effectiveness of techniques and technologies, applied in this research in automated loss estimation.

Topics & Keywords

3D Surveying and Cultural Heritage Conservation Techniques and Studies Cultural Heritage Materials Analysis

UN Sustainable Development Goals

Life in Land

Publication Details

Volume 20, pp. 390-394

DOI: 10.1109/atit50783.2020.9349254

Field-Weighted Citation Impact: 0.71