Scholarlink

<p>In the context of globalization and digitalization, there are both opportunities and challenges for integrating ethnic patterns into packaging design, and traditional methods are inefficient and poor in creative integration due to the lack of advanced algorithms. In this study, the EfficientNet image recognition and transfer learning algorithms were used to construct a new packaging design system, and the accuracy of ethnic pattern recognition on a large-scale dataset exceeded 95%, far exceeding that of traditional algorithms. The system analyzes design needs, selects fusion patterns, and generates packaging designs with a modern aesthetic. Through transfer learning to achieve cross-cultural design innovation, the experimental display system can improve design efficiency by 30% while maintaining design innovation and cultural depth, effectively shortening the design cycle. This shows that the intelligent design system based on EfficientNet significantly improves the application effect of ethnic pattern packaging design, high recognition accuracy ensures accurate pattern application, cross-cultural innovation gives packaging culture inclusiveness, and efficiency improvement strongly supports the application of ethnic patterns in the field of modern design, promotes the development of national culture communication, and explores more technology integration in the future to continuously optimize system performance.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Research on national pattern innovation fusion and packaging design system based on efficient net image recognition and transfer learning algorithm

  • 期刊论文

  • 发表日期：

    2025年4月1日

查看全文
  • Weijing Yuan,
  • Yinglin Ge

摘要

In the context of globalization and digitalization, there are both opportunities and challenges for integrating ethnic patterns into packaging design, and traditional methods are inefficient and poor in creative integration due to the lack of advanced algorithms. In this study, the EfficientNet image recognition and transfer learning algorithms were used to construct a new packaging design system, and the accuracy of ethnic pattern recognition on a large-scale dataset exceeded 95%, far exceeding that of traditional algorithms. The system analyzes design needs, selects fusion patterns, and generates packaging designs with a modern aesthetic. Through transfer learning to achieve cross-cultural design innovation, the experimental display system can improve design efficiency by 30% while maintaining design innovation and cultural depth, effectively shortening the design cycle. This shows that the intelligent design system based on EfficientNet significantly improves the application effect of ethnic pattern packaging design, high recognition accuracy ensures accurate pattern application, cross-cultural innovation gives packaging culture inclusiveness, and efficiency improvement strongly supports the application of ethnic patterns in the field of modern design, promotes the development of national culture communication, and explores more technology integration in the future to continuously optimize system performance.