Scholarlink

<p>This study explores the application of Artificial Intelligence Generated Content (AIGC) to the digital reconstruction of Qing-period interior canopy components, a representative form of traditional Chinese timberwork. To address fragmented archives and modeling inefficiency, we propose an integrated workflow combining historical image digitization, semantic lexicon building, prompt design, and image-to-model validation. A 76-term canopy glossary was embedded into a four-layer prompting template, and 312 images were generated across multiple platforms to analyze semantic response patterns and structural deviations. Validation using SketchUp confirmed stylistic fidelity with geometric deviations of only 3–4% and approximately 85% node interpretability. While AIGC excels in stylistic coherence and decorative richness, its limitations in structural logic require expert semantic annotation. Beyond reconstruction, the approach demonstrates potential for developing component ontologies, enriching BIM libraries, and supporting digital heritage conservation through rapid prototyping and stylistic diversity.</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

AIGC based digital heritage reconstruction of Qing interior canopies

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月30日

查看全文
  • Changqing Wei,
  • Dongyi Kong,
  • Yang Wang,
  • Jing Jia,
  • Jiaru Liu,
  • Yan Wei

摘要

This study explores the application of Artificial Intelligence Generated Content (AIGC) to the digital reconstruction of Qing-period interior canopy components, a representative form of traditional Chinese timberwork. To address fragmented archives and modeling inefficiency, we propose an integrated workflow combining historical image digitization, semantic lexicon building, prompt design, and image-to-model validation. A 76-term canopy glossary was embedded into a four-layer prompting template, and 312 images were generated across multiple platforms to analyze semantic response patterns and structural deviations. Validation using SketchUp confirmed stylistic fidelity with geometric deviations of only 3–4% and approximately 85% node interpretability. While AIGC excels in stylistic coherence and decorative richness, its limitations in structural logic require expert semantic annotation. Beyond reconstruction, the approach demonstrates potential for developing component ontologies, enriching BIM libraries, and supporting digital heritage conservation through rapid prototyping and stylistic diversity.