基于FMM和网络分析的唐代服饰纹样色彩复原与重用  

Restoration and reuse of clothing patterns and colors based on FMM and network analysis

作　　者：燕耀 张旭升 蔡欣华 雷桐 连璐 张寒诺 YAN Yao;ZHANG Xusheng;CAI Xinhua;LEI Tong;LIAN Lu;ZHANG Hannuo(School of New Media Art,Xi’an Polytechnic University,Xi’an 710048,China;Apparel&Art Design College,Xi’an Polytechnic University,Xi’an 710048,China;School of Computer Science and Engineering,Xi’an University of Technology,Xi’an 710048,China)

机构地区：[1]西安工程大学新媒体艺术学院,西安710048 [2]西安工程大学服装与艺术设计学院,西安710048 [3]西安理工大学计算机科学与工程学院,西安710048

出　　处：《丝绸》2024年第9期28-38,共11页Journal of Silk

基　　金：陕西省社会科学基金项目(2022J023)。

摘　　要：唐代服饰纹样作为中国古代服饰设计精髓,其绚丽丰富的色彩为当代服饰设计提供了宝贵的灵感和参考。为更准确地厘清唐代服饰纹样的用色偏好和配色规律,方便设计师进行唐代色彩重用,文章以唐代服饰文物遗存图像作为研究对象,提出了一种基于FMM和网络分析的唐代服饰纹样色彩复原与重用方法。首先,利用色彩分割算法在Lab空间对图像进行色彩分类分割,并基于FMM图像修复算法对受损及褪色区域进行修复和复原;其次,通过K-means聚类算法对图像色彩聚类计算,获得了图像色彩节点度和邻接关系,从而探明了唐代服饰纹样显性的用色、配色知识;再次,结合隐性配色知识,使用专家评价法创建了节点影响力效率矩阵,实现了基础色彩网络的优化和最终色彩模型的构建;最终,以丝巾产品作为应用案例,验证了所提方法的可行性和价值。案例验证表明,所提方法能够辅助设计师进行产品配色设计,并有效重用唐代服饰纹样的色彩设计知识。The Tang Dynasty,as one of the most prosperous eras in ancient China in terms of economy and culture,had distinctive and vibrant clothing patterns and color designs.These designs encapsulate the essence and cultural representation of Tang Dynasty attire,providing valuable creative inspiration and reference for contemporary fashion design.Currently,research on the color of clothing patterns of the Tang Dynasty mainly involves traditional subjective qualitative analysis and a small amount of quantitative color analysis.Quantitative analysis is primarily focused on analyzing and applying color samples of clothing patterns that have faded,been damaged,or polluted,which can lead to inaccuracies and biases in color analysis due to the deteriorated samples.Therefore,there is a need for image restoration and color restoration of the existing Tang Dynasty clothing patterns.At present,image restoration algorithms are mainly divided into traditional image restoration algorithms and deep learning image restoration algorithms.Among them,deep learning image restoration algorithms rely on the processing of a large amount of data and resources.Considering the scarcity of preserved images of Tang Dynasty clothing patterns,deep learning algorithms are not suitable.Therefore,the traditional image restoration algorithm FMM(fast marching method),which is technologically mature,is chosen.This algorithm does not depend on large-scale data computation and can quickly and accurately restore damaged areas.At the same time,due to its characteristic of repairing from the boundary inward,it can better cope with the main narrow and long damaged areas in the disease of Tang Dynasty clothing patterns while maintaining structural integrity and high clarity.Therefore,this algorithm is selected for the restoration of preserved images of Tang Dynasty clothing patterns.To more accurately clarify the color preferences and matching rules of Tang Dynasty clothing patterns,this study takes the preserved images of Tang Dynasty clothing cultural relics as

关 键 词：唐代服饰纹样 FMM算法 K-MEANS聚类 色彩复原 色彩提取 色彩网络模型 

分 类 号：TS941.26[轻工技术与工程—服装设计与工程]