Accès libre

Study on the Digital Restoration of Costumes in the Picture of Palace Ladies with Silk Fans in Tang Dynasty

Liuyun Zhang

Liuyun Zhang

Department of Digital Media Art,College of Architecture and Arts, Taiyuan University of TechnologyJinzhong, China
Recherchez cet auteur sur
Sciendo|Google Scholar
Zhang, Liuyun
 et 
Bo Xia

Bo Xia

Department of Digital Media Art,College of Architecture and Arts, Taiyuan University of TechnologyJinzhong, China
Recherchez cet auteur sur
Sciendo|Google Scholar
Xia, Bo
  
27 févr. 2025
Study on the Digital Restoration of Costumes in the Picture of Palace Ladies with Silk Fans in Tang Dynasty's Cover Image
À propos de cet article
Article précédent
Article suivant
Citez
Télécharger la couverture
Publié en ligne: 27 févr. 2025
Reçu: 12 oct. 2024
Accepté: 13 janv. 2025
DOI: https://doi.org/10.2478/amns-2025-0115
Mots clés
© 2025 Liuyun Zhang et al., published by Sciendo
This work is licensed under the Creative Commons Attribution 4.0 International License.

Figure 1.

Flowchart of image segmentation algorithm
Flowchart of image segmentation algorithm

Figure 2.

Image restoration technology architecture
Image restoration technology architecture

Figure 3.

K-Means algorithm initializes GMM clustering when k = 2
K-Means algorithm initializes GMM clustering when k = 2

Figure 4.

K-Means algorithm initializes GMM clustering when k = 6
K-Means algorithm initializes GMM clustering when k = 6

Figure 5.

Registration results of feature points
Registration results of feature points

Figure 6.

Training loss and validation loss of the model
Training loss and validation loss of the model

Figure 7.

Objective evaluation
Objective evaluation

Figure 8.

SSIM and PSNR values between original images and different restoration results
SSIM and PSNR values between original images and different restoration results

Figure 9.

Evaluation indexes of each model
Evaluation indexes of each model

Figure 10.

Effect of missing rate on algorithm performance
Effect of missing rate on algorithm performance

Figure 11.

Repair effect of different deletion rates
Repair effect of different deletion rates

Figure 12.

Objective evaluation index of different algorithms
Objective evaluation index of different algorithms

Error comparison of image matching results

ALGORITHM RMSE MAE MEE
CPD 86.1284 108.0754 28.6465
Ours 84.4054 106.0039 27.9689

Accuracy of Clothing Classification and Recognition in "Picture of a Lady with a Fan"

Models Original apparel recognition accuracy (%) Recognition accuracy of repaired clothing (%)
Vggnet _ v1 78.24 76.39
VGGNet _ v2 75.04 74.52
AlexNet 88.18 87.64
ResNet 94.72 94.18
Langue:
Anglais
Périodicité:
1 fois par an
Sujets de la revue:
Sciences de la vie, Sciences de la vie, autres, Mathématiques, Mathématiques appliquées, Mathématiques générales, Physique, Physique, autres
RSS Feed de la revue