Paper Title :Integrating Object Detection and Deep Learning for Oil Painting Paint Layer Defect Detection
Author :Yu-Ting Wu, Chwen-Tzeng Su, I-Cheng Li
Article Citation :Yu-Ting Wu ,Chwen-Tzeng Su ,I-Cheng Li ,
(2024 ) " Integrating Object Detection and Deep Learning for Oil Painting Paint Layer Defect Detection " ,
International Journal of Advance Computational Engineering and Networking (IJACEN) ,
pp. 12-17,
Volume-12,Issue-8
Abstract : The application of machine learning, deep learning, object detection and other related technologies has gradually
matured, but in the field of cultural asset protection and art restoration, related applications are just about to begin. In the
process of reviewing the restoration status of the works, in the past, it was necessary for the restorer to manually mark the
missing position, which resulted in a lot of time and energy consumption, and sometimes there was a possibility of
misjudgment due to the restorer's experience. To solve those problems, this study intends to use object detection technology
to assist restorers to speed up the efficiency of missing annotations in the work inspection process. This study proposes to
use YOLOv5 object detection technology to detect defects in the painted layer of oil paintings, assist restorers to save time in
manually marking damaged areas, and devote more time and energy to painting restoration operations. This study uses two
non-destructive examination techniques, Ultraviolet and Normal light, to identify retouching, insect excrement, and painted
layers loss. hoped this will give some technical assistance to the field of art restoration. Some repair techniques are combined
with today's computer technology to improve the effect of repair and reduce the consumption of manpower and time.
Keywords - Art Restoration, Defect Detection, Oil Painting Paint Layer Defect, YOLOv5
Type : Research paper
Published : Volume-12,Issue-8
DOIONLINE NO - IJACEN-IRAJ-DOIONLINE-21170
View Here
Copyright: © Institute of Research and Journals
|
 |
| |
 |
PDF |
| |
Viewed - 22 |
| |
Published on 2024-11-19 |
|