摘要
目的 针对青花瓷纹样物质文化遗产在目前所遭遇的发展困境以及人工智能迅速迭代发展的时代背景,通过深度学习技术识别纹样中的显性及隐性信息,实现对纹样的语义解读,以将其应用于个性化的产品设计。方法 以明清青花瓷瓶纹样为研究样本,从本体和隐义两个方面对图像进行标签化处理,构建图像双层多标签标注模型,实现对陶瓷饰纹的寓意解读。结果设计了明清青花瓷瓶纹样交互式文创产品定制App。结论 将深度学习的图像识别技术有应用于青花瓷纹样可以传承和发扬我国优秀传统陶瓷文化。
Objective In view of the development difficulties of the material cultural heritage of blue and white porcelain patterns and the rapid iterative development of artificial intelligence,the study explores the explicit and implicit information in the patterns through deep learning technology to realize the symbolic interpretation of the patterns, so as to apply them to the design of personalized products. Methods Taking the Ming and Qing dynasty blue and white porcelain vase patterns as research samples, the images are labeled from both ontological and implicit meanings, and the image two-layer multi-label annotation model is constructed to achieve the semantic interpretation of ceramic decorative patterns. Results An interactive cultural and creative product customization app for Ming and Qing dynasty blue and white vase patterns is designed. Conclusion The application of deep learning image recognition technology to porcelain patterns can inherit and carry forward China's excellent traditional ceramic culture.
出处
《包装与设计》
2024年第3期100-101,共2页
Package & Design
关键词
深度学习
语义识别
青花瓷纹样
个性化设计
多标签标注模型
deep learning
semantic recognition
blue and white porcelain pattern
personalized design
multi-label annotation models