基于蝴蝶呈色风格的毛衫花型设色迁移研究
Study of Pattern Coloration of Knitted Sweaters Based on Butterfly Color Style
蒋汶秦 1闫竞文 1罗芊芊 1徐平华2
作者信息
- 1. 浙江理工大学服装学院,浙江杭州 310018
- 2. 浙江理工大学服装学院,浙江杭州 310018;浙江理工大学浙江省哲学社会科学重点培育研究基地数智风格与创意设计研究中心,浙江杭州 310018;浙江理工大学丝绸文化传承与产品设计数字化技术文化和旅游部重点实验室,浙江杭州 310018
- 折叠
摘要
文章选用3种不同色彩迁移算法,将蝴蝶的配色风格自动迁移至目标纹样中.将输出纹样与针织毛衫设计相结合,采用虚拟设计的形式呈现不同种类蝴蝶呈色风格.试验选取5类典型蝴蝶作为色彩参考源,对Reinhard、Welsh和Fast Neural Style(FNS)3种算法迁移效果进行对比发现,FNS算法生成图像色彩过渡自然、底色未受影响,与参考图像风格接近,色彩迁移平均耗时每张图片0.37 s.该研究从蝴蝶图像中提炼色彩特征,为针织毛衫设计提供了新思路,对其他类型的服饰产品设计具有一定参考价值.
Abstract
In this study,three different color transfer algorithms were used to automatically transfer the color styles of butterflies to the target pattern.The output patterns were combined with the design of knitted sweaters,and different types of butterfly color styles were presented in the form of virtual design.Five typical butterfly species were selected as color reference sources,and the transfer effects of three algorithms including Reinhard,Welsh and Fast Neural Style(FNS)were compared.The results show that the FNS algorithm generated natural color transitions and the background color is not affected,which is similar to the style of the reference image,and the average time-cost for color transfer is 0.37 s/pic per image.This study extracted color characteristics from butterfly images,providing new ideas for the design of knitted sweaters with reference value for the design of other types of fashion products.
关键词
蝴蝶色彩/色彩迁移/针织毛衫/配色设计/纹理映射Key words
Butterfly Color/Color Transfer/Knitted Sweater/Coloration Design/Texture Mapping引用本文复制引用
基金项目
浙江省哲学社会科学规划交叉学科及冷门"绝学"课题(24LMJX09YB)
浙江省重点研发计划(2024C01210)
浙江省研究生教育学会科研项目(2023-012)
浙江理工大学研究生教育教学改革研究项目(YJG-Z202301)
浙江理工大学基本科研业务费项目(24076109-Y)
浙江省大学生科技创新活动计划(新苗人才计划)项目(2024)(77)
国家级大学生创新创业训练计划(202310338047)
出版年
2024
NETL
NSTL
万方数据