Research on Image Feature Analysis of Intangible Cultural Heritage Brocade Integrating Multi-Scale Visual Perception

Authors

• Ruiyang YUAN, 1.School of Information Management, Nanjing University, Nanjing 210023 2.Jiangsu Key Laboratory of Data Engineering and Knowledge Service, Nanjing 210023
• Hao WANG, 1.School of Information Management, Nanjing University, Nanjing 210023 2.Jiangsu Key Laboratory of Data Engineering and Knowledge Service, Nanjing 210023
• Shu ZHOU, 1.School of Information Management, Nanjing University, Nanjing 210023 2.Jiangsu Key Laboratory of Data Engineering and Knowledge Service, Nanjing 210023
• Hui ZHU, 1.School of Information Management, Nanjing University, Nanjing 210023 2.Jiangsu Key Laboratory of Data Engineering and Knowledge Service, Nanjing 210023
• Jingwen QIU, 1.School of Information Management, Nanjing University, Nanjing 210023 2.Jiangsu Key Laboratory of Data Engineering and Knowledge Service, Nanjing 210023

Keywords

intangible cultural heritage, digital humanities, image feature analysis, feature interpretability, self-attention mechanism

Abstract

[Purpose/significance] Addressing the challenges posed by the complex semantic characteristics of intangible cultural heritage brocade imagery, the difficulty in extracting their profound connotations, and the inadequate utilisation of multi-scale features by traditional deep learning models, this paper aims to explore a method for analysing the characteristics of intangible cultural heritage brocade images that integrates multi-scale visual features. [Method/process] This paper constructs a multi-scale feature analysis framework for intangible cultural heritage brocade images (ICH_BC), integrating convolutional neural networks with Transformer architectures. The framework employs ResNet to extract local texture and detail features from brocade images, utilises VIT to capture global structural and stylistic information, and dynamically weight-combines features across scales through self-attention mechanisms to achieve collaborative modelling of multi-scale visual features. Building upon this foundation, the framework employs UMAP dimensionality reduction for visualisation alongside Class Activation Maps (CAM) to analyse the high-dimensional feature distributions and regions of interest learned by the model. This facilitates understanding of the model's discriminative principles and explores pathways for translating digital features into cultural characteristics. [Result/conclusion] The experiment validated the framework using eight categories of nationally designated intangible cultural heritage brocade images. Results demonstrated that the ICH_BC framework outperformed single-scale feature models and static feature fusion methods in classification performance, confirming the efficacy of multi-scale visual feature fusion. Feature visualisation revealed distinct spatial distributions across different brocade categories, indicating varying model emphasis on local patterns versus overall compositional features. This approach provides an effective and generalizable technical pathway and analytical paradigm for understanding and analysing cultural heritage images with complex visual characteristics. It holds theoretical and practical value for advancing the digital research, conservation, and dissemination of cultural heritage.

First Page

114

Last Page

125

Submission Date

24-Nov-2025

Revision Date

27-Jan-2026

Acceptance Date

10-Jun-2026

Published Date

1-Jul-2026

Reference

[1] 刘炜，叶鹰.数字人文的技术体系与理论结构探讨[J].中国图书馆学报，2017，43（5）：32-41.

[2] 张洋.数字化时代科技赋能非遗文化传承与推广的路径探究[J].教育传媒研究，2024（5）：93-95.

[3] 张佳南，于洪，王延飞，等.情报视角下新质生产力的智能变革[J].科技情报研究，2024，6（4）：1-10.

[4] 王兰伟.新质生产力赋能公共文化服务均等化的创新思路与实践路径研究[J].图书馆建设，2024（4）：74-82.

[5] 刘鸿达，孙旭辉，李沂滨，等.基于卷积神经网络的图像分类深度学习模型综述[J].计算机工程与应用，2025，61（11）：1-21.

[6] DOSOVITSKIY A，BEYER L，KOLESNIKOV A，et al.An Image is Worth 16×16 Words：Transformers for Image Recognition at Scale[J].arXiv e-prints，arXiv：2010.11929.

[7] VASWANI A，SHAZEER N，PARMAR N，et al.Attention is all you need[J].Advances in Neural Information Processing Systems，2017，30：6000-6010.

[8] BURRELL J.How the machine 'thinks'：Understanding opacity in machine learning algorithms[J].Big Data & Society，2016，3（1）：1-12.

[9] 黄永林，谈国新.中国非物质文化遗产数字化保护与开发研究[J].华中师范大学学报（人文社会科学版），2012，51（2）：49-55.

[10] 董坤.基于知识元的非物质文化遗产知识抽取与组织研究[J].情报理论与实践，2021，44（9）：155-160，148.

[11] 范青，史中超，谈国新.非物质文化遗产的知识图谱构建[J].图书馆论坛，2021，41（10）：100-109.

[12] DOU J H，QIN J Y，JIN Z X，et al.Knowledge graph based on domain ontology and natural language processing technology for Chinese intangible cultural heritage[J].Journal of Visual Languages & Computing，2018，48：19-28.

[13] 陈昱成，黎洋，刘江峰，等.AIGC视角下非物质文化遗产知识图谱的构建研究[J].科技情报研究，2024，6（2）：115-128.

[14] 陈娟，王诗晓，刘舒娴.数字工具促进织锦非遗工艺更迭研究[J].丝绸，2025，62（9）：1-10.

[15] LOWE D G.Distinctive Image Features from Scale-Invariant Keypoints[J].International Journal of Computer Vision，2004，60（2）：91-110.

[16] KRIZHEVSKY A，SUTSKEVER I，HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM，2017，60（6）：84-90.

[17] SIMONYAN K，ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[J].arXiv e-prints，arXiv：1409.1556.

[18] HE K M，ZHANG X Y，REN S Q，et al.Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Las Vegas，2016：770-778.

[19] LIN T-Y，DOLLÁR P，GIRSHICK R，et al.Feature Pyramid Networks for Object Detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Honolulu，HI，2017：936-944.

[20] BAI L，LIU Q X，LI C L，et al.Remote Sensing Image Scene Classification Using Multiscale Feature Fusion Covariance Network with Octave Convolution[J].IEEE Transactions on Geoscience and Remote Sensing，2022，60：1-14.

[21] 陈佳慧，路鹏，罗小玲，等.基于CNN与ViT混合结构的遥感图像地物分类算法[J].遥感信息，2024，39（3）：121-127.

[22] FIORUCCI M，KHOROSHILTSEVA M，PONTIL M，et al.Machine Learning for Cultural Heritage：A Survey[J].Pattern Recognition Letters，2020，133：102-108.

[23] 周泽聿，王昊，张小琴，等.基于Xception-TD的中华传统刺绣分类模型构建[J].数据分析与知识发现，2022，6（Z1）：338-347.

[24] 石斌，王昊，邓三鸿.基于图像模态迁移与集成的中国陶瓷派系识别模型构建研究[J].数据分析与知识发现，2023，7（12）：88-101.

[25] 范涛，王昊，李跃艳，等.基于多模态融合的非遗图片分类研究[J].数据分析与知识发现，2022，6（Z1）：329-337.

[26] SELVARAJU R R，COGSWELL M，DAS A，et al.Grad-CAM：Visual Explanations from Deep Networks via Gradient-based Localization[C]//Proceedings of the IEEE International Conference on Computer Vision.Venice，2017：618-626.

[27] MCINNES L，HEALY J，SAUL N，et al.UMAP：Uniform Manifold Approximation and Projection for Dimension Reduction[J].Journal of Open Source Software，2018，3（29）：1-63.

[28] SHAMIR L，TARAKHOVSKY J A.Computer analysis of art[J].Journal on Computing and Cultural Heritage，2012，5（2）：7:1-7:11.

[29] 郭佩林，张德，王怀秀.基于特征可视化探究跳跃连接结构对深度神经网络特征提取的影响[J].计算机工程，2025，51（4）：149-157.

[30] 盛家川，陈雅琦，韩亚洪.深层网络特征聚合重标定的中国画情感分类算法[J].计算机辅助设计与图形学学报，2020，32（9）：1420-1429.

[31] 周飞燕，金林鹏，董军.卷积神经网络研究综述[J].计算机学报，2017，40（6）：1229-1251.

[32] 司念文，张文林，屈丹，等.卷积神经网络表征可视化研究综述[J].自动化学报，2022，48（8）：1890-1920.

[33] ZHOU B，KHOSLA A，LAPEDRIZA A，et al.Learning Deep Features for Discriminative Localization[C]//Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition.New York，2016：2921-2929.

[34] 钱小萍.蜀锦、宋锦和云锦的特点剖析[J].丝绸，2011，48（5）：1-6.

[35] 王宝林.南京云锦的文化内涵[J].学海，2003（2）：167-170.

[36] 任雪玲，葛玉珍.鲁锦的艺术特色及基础纹样解析[J].丝绸，2009（6）：45-48.

[37] 叶洪光，冯泽民.土家织锦的文化生态与视觉艺术[M].武汉：武汉理工大学出版社，2016.

[38] 周赳.中国古代三大名锦的品种梳理及美学特征分析[J].丝绸，2018，55（4）：93-105.

[39] 周赳，吴文正.中国古代织锦的技术特征和艺术特征[J].纺织学报，2008（3）：47-50.

[40] BECHT E，MCINNES L，HEALY J，et al.Dimensionality reduction for visualizing single-cell data using UMAP[J].Nature Biotechnology，2018，37：38-44.

[41] 徐杰舜.文化基因：五论中华民族从多元走向一体[J].湖北民族学院学报（哲学社会科学版），2008（3）：9-14.

[42] 陈文雯，刘春芳，龚小凡.浅谈唐代蜀锦的审美思想[J].艺术与设计（理论），2008（7）：232-234.

[43] 蔡荭.广西壮锦、瑶锦、苗锦、侗锦传统纹样探析[J].广西文博，2017（00）：334-353.

Digital Object Identifier (DOI)

10.19809/j.cnki.kjqbyj.2026.03.012

Recommended Citation

YUAN, Ruiyang; WANG, Hao; ZHOU, Shu; ZHU, Hui; and QIU, Jingwen (2026) "Research on Image Feature Analysis of Intangible Cultural Heritage Brocade Integrating Multi-Scale Visual Perception," Journal of Scientific Information Research: Vol. 8: Iss. 3, Article 12.
DOI: 10.19809/j.cnki.kjqbyj.2026.03.012
Available at: https://eng.kjqbyj.com/journal/vol8/iss3/12