首页|期刊导航|鞋类工艺与设计|纺织品图案设计中人工智能生成与手绘融合路径研究——以纺织品四方连续图案为例

纺织品图案设计中人工智能生成与手绘融合路径研究——以纺织品四方连续图案为例OA

Research on the Integration Pathways of AI-Generation and Hand-Drawing Patterns in Textile Design——Taking Textile Square Continuous Patterns as an Example

中文摘要英文摘要

本文旨在探讨人工智能与手绘在纺织品图案设计中的融合可行性,以四方连续图案为研究对象,分析手绘与人工智能生成的优势与短板,进而形成"人机共创"的融合流程.研究选取敦煌西魏壁画为元素来源,分别开展人工智能与手绘设计实践.手绘实践中采用传统绘制结合Photoshop完成图案设计;人工智能生成实践中,利用ChatGPT-5 按照文本描述生成图案.对比实验结果表明,手绘图案元素组合留白较少,主元素呈曲线形循环;AI生成图案元素留白较多,主元素为直线或阶梯式循环,需进一步人工修正.本研究提出"手绘―图像生成型AI生成―融合反馈"路径,为纺织品设计思维与教育提供全新思路.

This paper aims to explore the feasibility of integrating artificial intelligence and hand-drawing in textile pattern design,using four-way repeat patterns as the research object.It analyzes the strengths and weaknesses of hand-drawing and AI generation,and then develops a"human-AI co-creation"workflow.The study takes Dunhuang Western Wei murals as the source of design elements and carries out two types of design practice.In the hand-drawing practice,traditional drawing methods combined with Photoshop are used to complete the pattern design.In the AI generation practice,ChatGPT-5 is used to generate patterns based on text descriptions.The comparison shows that hand-drawn patterns have less blank space and the main elements form curved repetition,while AI-generated patterns contain more blank space and the main elements appear in straight or step-like repetition,so further manual revision is needed.This study proposes a"hand-drawing ― image-generating AI ― integration and feedback"path,providing new ideas for textile design thinking and education.

史泽萱;刘德川

韩国成均馆大学,韩国 首尔 03063韩国东西大学设计学院,韩国 釜山 47011

社会科学

纺织品图案设计人工智能生成手绘四方连续图案融合路径

textile pattern designai generationhand-drawingtextile square continuous patternintegration pathways

《鞋类工艺与设计》 2026 (19)

11-13,3

10.3969/j.issn.2096-3793.2026-19-004

评论