With the continuous development of science, combining art and science has become a new means of conservation. This paper first analyzes the application of deep learning in the field of NRM conservation, and then implements paper cutting pattern design in three aspects: generating fractal patterns, network architecture design, and perceptual loss function. Finally, this paper proposes a strategy of digital conservation of paper cutting non-heritage based on deep learning by cultivating intelligent interpreters, constructing a paper cutting database, and reconstructing a paper cutting virtual field to develop ideas for designing paper cutting patterns as well as spreading, protecting, and passing on paper cutting non-heritage.

Lulu Zhao. Research on the Design of Paper Cutting Patterns and Digital Preservation Strategy of Non-heritage Based on Deep Learning. Frontiers in Art Research (2022) Vol. 4, Issue 15: 91-94. https://doi.org/10.25236/FAR.2022.041518.

[1] Zhu Yinghui, Jiang Yuzhen. An algorithm for generating stylized paper-cut images by deep learning [J]. Journal of Hanshan Normal University, 2022, 43(03): 21-28+35. doi: 10.19986/j. cnki. 1007- 6883.2022.03. 

[2] Chen Yuhong, Liu Xiaojing. Research on automatic classification of Thangka statues based on convolutional neural networks [J]. Computer Technology and Development, 2021, 31 (12): 167-174.

[3] Yao Chao, Liu Guihua, Zhao Sen, Lin Jie. Paper-cutting design of characters based on convolutional neural network [J]. Sensors and Microsystems, 2021, 40 (07): 88-91. doi: 10.13873/J.1000-9787 (2021) 07-0088 

[4] Nie Wei. Practice and Development of Intelligent Communication of Intangible Cultural Heritage [J]. Sichuan Drama, 2019(11):97-101. 

[5] Huang Changzheng, Chen Xi, Zhou Yanming. Application and design realization of artificial intelligence deep learning algorithm in virtual reality interactive products [J]. Research on Science and Technology Innovation Development Strategy, 2019, 3 (01): 16-24.