纺织学报 ›› 2024, Vol. 45 ›› Issue (06): 155-164.doi: 10.13475/j.fzxb.20230705901

• 服装工程 • 上一篇    下一篇

《宫乐图》中仕女服饰的数字化复原

黄智威, 石雅杰, 陆冯, 李青, 薛哲彬()   

  1. 苏州大学 纺织与服装工程学院, 江苏 苏州 215021
  • 收稿日期:2023-07-25 修回日期:2023-10-21 出版日期:2024-06-15 发布日期:2024-06-15
  • 通讯作者: 薛哲彬(1984—),女,教授,博士。主要研究方向为多模态时尚感知挖掘及应用,数字化时尚。E-mail: zhebin.xue@suda.edu.cn
  • 作者简介:黄智威(1998—),男,硕士生。主要研究方向为数字化服装设计。
  • 基金资助:
    教育部人文社科青年基金项目(22YJC760111);江苏省丝绸工程重点实验室开放课题(KJS2057)

Digital restoration of women's costumes depicted in ancient painting of "A Palace Concert"

HUANG Zhiwei, SHI Yajie, LU Feng, LI Qing, XUE Zhebin()   

  1. College of Textile and Clothing Engineering, Soochow University, Suzhou, Jiangsu 215021, China
  • Received:2023-07-25 Revised:2023-10-21 Published:2024-06-15 Online:2024-06-15

摘要:

通过数字化手段对古代画作中的人物服饰进行重现可达到简化传统服饰复原流程、节约修复成本等目的。以《宫乐图》中端茶仕女为例探索古代服饰的数字化复原方法论。首先,借助数字化服装制图技术对原画受损部分进行修复,并基于文史资料及服装美学原理,对仕女服款式、结构等信息进行分析和推断;而后,通过数字化制板技术绘制服装样板,运用三维虚拟现实技术对服饰进行数字化还原展示;最后,结合层次分析法及模糊综合评价法构建三层级的评价体系对复原效果进行验证。结果表明,复原后的仕女服不论从服装造型、结构、面料及工艺的还原程度,还是服饰的整体复原效果,其好评占比均达到40%以上,说明本文提出的以古代画作为参考进行的古代服饰数字化复原方法可行,效果满意。

关键词: 宫乐图, 仕女服饰, 数字化复原, 虚拟展示, 多层次模糊综合评价

Abstract:

Objective The traditional process of restoring ancient costumes relies on physical artifacts as a basis, which often leads to cumbersome and time-consuming restoration procedures due to factors such as the condition of cultural relics, restoration processes, and the experience of researchers. The restoration of ancient costumes necessitates broader exploration and the quest for a more scientific methodology to serve as guidance.

Method Taking the tea-serving maiden in the ancient painting of "A Palace Concert" as the subject, a digital approach was adopted to restore the Tang Dynasty women's costumes depicted in the painting. This approach involved painting restoration using digital techniques, inferencing costume styles, sizes and other information through the use of literature and aesthetic principles, sample production using digital pattern-making tools, and virtual costume creation and display using three-dimensional virtual-reality technology. The evaluation system was constructed using the Analytic Hierarchy Process (AHP) and consists of four primary criteria and eleven secondary criteria. The weights of each criterion were calculated, and the restoration results were evaluated using the fuzzy comprehensive evaluation method.

Result A comprehensive evaluation system and criteria for the restoration of women's costumes depicted in the Painting "A Palace Concert" were constructed, including four primary criteria, which were "Overall Styling", "Costume Construction", "Fabric Selection", and "Craftsmanship", as well as eleven secondary criteria to assess the effectiveness of costume restoration. Using the expert assessment method, weights were obtained for both the primary and secondary criteria. The primary criteria assigned 0.243 3 weight to "Overall Styling", 0.378 9 to "Costume Construction", 0.162 3 to "Fabric Selection", and 0.188 0 to "Craftsmanship", among which, "Costume Construction" had the highest weight. The secondary criteria assigned "Costume Bodice" and "Costume Style" with relatively high weights, with values of 0.59 and 0.488 7, respectively. Their comprehensive weights were 0.143 5 and 0.185 2 respectively, indicating that the restoration effectiveness of the costume bodice and style silhouette, which are the main visual perception elements, significantly influence the final restoration outcome of the women's costumes. The virtual restoration results were evaluated for the secondary criteria, and the evaluation degrees of "Very Good", "Good", "Fair", "Poor", and "Very Poor" were obtained for corresponding elements. The evaluation results for the secondary criteria were then integrated and statistically analyzed. It can be observed that all secondary criteria received evaluation degrees of "Very Good" greater than 35%, and there were seven elements "Costume Silhouette", "Costume Sleeves", "Costume Collar", "Fabric Material", "Fabric Color", "Fabric Pattern" and "Piece Seaming" with evaluation degrees greater than 40%. This indicates that the restoration effectiveness for all secondary criteria was rated as "Very Good". Furthermore, the evaluation was recalculated with consideration of weights for the primary indicators and overall restoration effectiveness. All primary indicators were rated as "Very Good", with the breakdown as follows:"Overall Styling" at 40.25%, "Costume Construction" at 41.85%, "Fabric Selection" at 47.68%, and "Craftsmanship" at 42.17%. The final evaluation result for the "Overall Restoration Outcome" was categorized as "Very Good" at 41.32%. These data demonstrated that the restoration of the primary criteria and the overall effect of the Qi-chest Ruqun costume for the tea-serving maiden in the Painting of "A Palace Concert" were highly successful.

Conclusion By using digital techniques for restoration and incorporating literature and other images, combined with the utilization of CLO 3D software, it is possible to achieve digital restoration of ancient women's costumes depicted in the Painting of "A Palace Concert" and provide a viable methodology for exploring the digital restoration of ancient costumes. In terms of evaluation, the "Overall Styling", "Costume Construction", "Fabric Selection", and "Craftsmanship" have varying degrees of impact on the authenticity of virtual costume restoration based on the painting. Among them, the "Costume Bodice" and "Costume Style" have a relatively higher influence, indicating that the overall visual presentation is the most critical factor influencing the effectiveness of costume digital restoration. The final restoration results received evaluations of "Very Good" and "Good," with a combined percentage of over 70%, demonstrating that digital restoration based on ancient paintings can effectively simulate and restore ancient costumes with good visual effects and feasibility.

Key words: painting of "A Palace Concert", women's costume, digital restoration, virtual display, multilevel fuzzy comprehensive evaluation

中图分类号: 

  • TS941

图1

《宫乐图》原画"

图2

AI修复《宫乐图》"

图3

PS修复的《宫乐图》"

图4

选择复原对象"

图5

还原款式图"

图6

纹饰复原结果"

表1

复原服饰颜色推断"

服饰 颜色 潘通色号
低饱和度、高明度的绿色 SM1-1191
裙、披帛 高饱和度、较高明度的红色 PANTONE 7628 C
系带 低饱和度、高明度的蓝色 COLORO 107-66-11
披帛纹饰 中饱和度、高明度的黄色 NCS S 2040-Y

表2

设置身体各部位数据"

身高 胸围 腰围 臀围 颈围 膝围 腕围 肩宽
160 84 66 88 34 32 16 38

图7

虚拟模特构建"

表3

尺寸推算"

服装 部位 与身高比例 尺寸/cm
衣长 1∶4 40
胸围 1∶1.7 96
袖长 1∶4 40
袖口宽 1∶8.3 20
裙长 1∶1.11 145
腰围 1∶1.23 130
下摆围 1∶0.62 260
披帛 1∶1.9 300
1∶16 10

图8

衣片制作"

图9

逐层建模过程"

图10

面料模拟"

图11

渲染结果展示"

表4

评价层级"

目标层 标准层 指标层
《宫乐图》仕女数字化复原效果B 整体造型b1 服装廓形b11
服装款式b12
服装构造b2 服装衣身b21
服装衣袖b22
服装领部b23
服装腋下b24
服装下摆b25
服装面料b3 面料材质b31
面料颜色b32
面料图案b33
制作工艺b4 布片拼缝b41

表5

判断维度"

标度 定义 内容
1 相同 两项指标重要程度相同
3 略微重要 一项指标较另一项略微重要
5 显著重要 一项指标较另一项显著重要
7 非常重要 一项指标较另一项非常重要
9 至关重要 一项指标较另一项至关重要
2、4、6、8 介于相邻标度之间
倒数 若因素i与因素j之间的重要性之比为aij,那么因素j与因素i重要性之比为aji=1/aij

表6

各层效果指标权重"

一级指标 权重 二级指标 权重 综合权重
整体造型 0.243 3 服装廓形 0.41 0.099 7
服装款式 0.59 0.143 5
服装构造 0.378 9 服装衣身 0.488 7 0.185 2
服装衣袖 0.136 8 0.051 8
服装领部 0.162 9 0.061 7
服装腋下 0.075 8 0.028 7
服装下摆 0.135 7 0.051 4
服装面料 0.162 3 面料材质 0.462 3 0.075 0
面料颜色 0.202 9 0.032 9
面料图案 0.334 7 0.054 3
制作工艺 0.188 0 布片拼缝 1 0.188 0

表7

虚拟复原效果评价统计权重"

标准层 指标层 很好 较好 一般 较差 很差
整体造型 服装廓形 0.409 6 0.313 2 0.096 3 0.096 3 0.084 3
服装款式 0.397 6 0.361 4 0.060 2 0.096 3 0.084 3
服装构造 服装衣身 0.397 6 0.349 4 0.072 3 0.048 2 0.132 5
服装衣袖 0.494 0 0.241 0 0.084 3 0.084 3 0.096 3
服装领部 0.481 9 0.228 9 0.108 4 0.096 3 0.084 3
服装腋下 0.361 4 0.409 6 0.048 2 0.108 4 0.071 4
服装下摆 0.373 5 0.385 5 0.060 2 0.060 2 0.120 5
服装面料 面料材质 0.494 0 0.277 1 0.048 2 0.108 4 0.072 3
面料颜色 0.409 6 0.349 4 0.060 2 0.144 6 0.036 1
面料图案 0.494 0 0.289 1 0.036 1 0.084 3 0.096 3
制作工艺 布片拼缝 0.421 7 0.325 3 0.072 3 0.072 3 0.108 4
[1] 李华飙, 孟竹, 李洋, 等. 古代服饰数字化保护与修复技术初探[C]//北京数字科普协会. 数字技术拓展博物馆服务:2021年北京数字博物馆研讨会论文集. 北京: 中国戏剧出版社,2021: 256-263.
LI Huabiao, MENG Zhu, LI Yang, et al. Exploring digital preservation and restoration techniques for ancient costumes[C]//Beijing Association for Digital Science and Technology Popularization. Digital Technology Expanding Museum Services:Proceedings of the 2021 Beijing Digital Museum Symposium. Beijing: China Drama Press,2021: 256-263.
[2] 姚彤, 黄楚懿. 基于CLO 3D平台的汉服数字化复原与创新设计研究[J]. 山东纺织科技, 2021, 62(4): 38-42.
YAO Tong, HUANG Chuyi. Research on digital restoration and innovative design of Hanfu based on ClO 3D platform[J]. Shandong Textile Science & Technology, 2021, 62(4): 38-42.
[3] 梁盈春, 宋晶鑫. 基于《九宫女图》的唐装三维仿真复原[J]. 纺织科技进展, 2022, 263(12): 43-46.
LIANG Yingchun, SONG Jingxin. 3D simulation restoration of Tang Dynasty costumes based on "Nine Palace Maids"[J]. Progress in Textile Science & Technology, 2022, 263(12): 43-46.
[4] 刘凯旋, 李晓宁, 朱春. 《簪花仕女图》仕女服饰复原研究[J]. 丝绸, 2022, 59(5): 59-67.
LIU Kaixuan, LI Xiaoning, ZHU Chun. Study on the restoration of ladies'costumes in Beauties Wearing Flowers[J]. Journal of Silk, 2022, 59(5): 59-67.
[5] 张瑜. 汉画像石《家宴图》人物服饰造型研究及3D仿真复原设计[D]. 天津: 天津工业大学, 2021: 1-4.
ZHANG Yu. Research on the styling of costume in Han portrait stone "Jiayan Tu" and 3D simulation restoration design[D]. Tianjin:Tiangong University, 2021: 1-4.
[6] 刘东升, 史慧, 刘运娟, 等. 乌珠穆沁部落蒙古袍形制结构研究及虚拟试衣[J]. 丝绸, 2020, 57(8): 50-56.
LIU Dongsheng, SHI Hui, LIU Yunjuan, et al. Study on shape structure of Mongolian robe of Ujimqin tribe and virtual fitting[J]. Journal of Silk, 2020, 57(8): 50-56.
[7] 杨永华, 张皋鹏. 基于虚拟仿真技术的羌族女装三维数字化展示研究[J]. 西部皮革, 2020, 42(11): 118-119.
YANG Yonghua, ZHANG Gaopeng. Research on the three-dimensional digital display of Qiang ethnic women's clothing based on virtual simulation tech-nology[J]. Western Leather, 2020, 42(11): 118-119.
[8] 王靖涵. 唐代仕女人物画服饰的文化内涵研究[J]. 美术教育研究, 2023, 291(8): 13-15.
WANG Jinghan. Research on the cultural connotation of clothing in Tang Dynasty figure paintings of women[J]. Art Education Research, 2023, 291(8): 13-15.
[9] 张弛. 唐代仕女画中的女性服装研究[D]. 扬州: 扬州大学, 2022: 1-47.
ZHANG Chi. Research on the female clothing of the beauty paintings of Tang Dynasty[D]. Yongzhou: Yangzhou University, 2022: 1-47.
[10] 沈从文. 中国古代服饰研究[M]. 上海: 上海书店出版社, 2005: 293-300.
SHEN Congwen. Research on ancient chinese costumes[M]. Shanghai: Shanghai Bookstore Publishing House, 2005: 293-300.
[11] 王勇. 《宫乐图》考[J]. 收藏, 2019, 364(10): 54-61.
WANG Yong. A study of "Painting Song and Music Performance at Palace"[J]. Collections, 2019, 364(10): 54-61.
[12] QIN Z, ZENG Q, ZONG Y, et al. Image inpainting based on deep learning: a review[J]. Displays, 2021. DOI:10.1016/j.displa.2021.102028.
[13] 刘拓, 赵忠波. 中国工笔重彩绘画数字化复原研究[J]. 美术教育研究, 2022, 269(10): 52-53.
LIU Tuo, ZHAO Zhongbo. Research on the digital restoration of Chinese gouache painting[J]. Art Education Research, 2022, 269(10): 52-53.
[14] KIM T, YEO B C. Recovering microscopic images in material science documents by image inpainting[J]. Applied Sciences, 2023. DOI: 10.3390/app13064071.
[15] 包铭新. 唐代女装的腰线及其审美效果[J]. 中国纺织大学学报, 1991(Z1): 1-11.
BAO Mingxin. The waistline women's costumes in Tang Dynasty and its aesthetic effects[J]. Journal of China Textile University, 1991(Z1): 1-11.
[16] 王欣然. 《宫乐图》中弹筝仕女服装复原[J]. 轻纺工业与技术, 2022, 51(5): 33-36.
WANG Xinran. Restoration of the costume of zither-playing women in "Gong Le Tu"[J]. Light and Textile Industry and Technology, 2022, 51(5): 33-36.
[17] 陶辉, 苏亚兰, 邹颖, 等. 隋唐时期襦裙服饰艺术特征及成因探析[J]. 服饰导刊, 2023, 12(1): 15-23.
TAO Hui, SU Yalan, ZOU Ying, et al. An analysis of the artistic characteristics and causes of the jacket dress in the Sui and Tang Dynasties[J]. Fashion Guide, 2023, 12(1): 15-23.
[18] 周讯, 高春明. 中国历代妇女妆饰[M]. 上海: 学林出版社, 1988: 218-223.
ZHOU Xun, GAO Chunming. Adornments of women in ancient China[M]. Shanghai: Shanghai People's Publishing House, 1988: 218-223.
[19] 李柠. 唐襦裙服饰形式美的数字化展示研究[D]. 西安: 西北大学, 2018: 10-16.
LI Ning. Digital display and research of formal beauty of skirt clothing of Tang Dynasty[D]. Xi'an: Northwest University, 2018: 10-16.
[20] 曾艳红. 唐诗丝绸物事及其意象研究[D]. 上海: 上海师范大学, 2010: 9-10, 29-32.
ZENG Yanhong. The study of silk things and its images in Tang Dynasty poetry[D]. Shanghai: Shanghai Normal University, 2010: 9-10, 29-32.
[21] 罗依坤. 唐宋女性服饰色彩研究[D]. 株洲: 湖南工业大学, 2016: 20-25, 30.
LUO Yikun. Study on the Tang and Song Dynasty women dress color[D]. Zhuzhou: Hunan University of Technology, 2016: 20-25, 30.
[22] 梁佩珊. 《唐人宫乐图》艺术特色研究[D]. 兰州: 西北师范大学, 2019: 51-53.
LIANG Peishan. A study on the artistic characteristics of the music picture of the Tang Dynasty palace[D]. Lanzhou: Northwest Normal University, 2019: 51-53.
[23] 孙蕾, 朱泓. 郑州地区汉唐宋成年居民的身高研究[J]. 人类学学报, 2015, 34(3): 377-389.
SUN Lei, ZHU Hong. A study of human statures of skeletons found in Han, Tang and Song Dynasty tombs, Zhengzhou area[J]. Acta Anthropologica Sinica, 2015, 34(3): 377-389.
[24] 武敏. 吐鲁番考古资料所见唐代妇女时装[J]. 西域研究, 1992(1): 62-75.
WU Min. Research on women's fashion in the Tang Dynasty based on archaeological materials from Turpan[J]. The Western Regions Studies, 1992(1): 62-75.
[25] 韩利, 梅强, 陆玉梅, 等. AHP-模糊综合评价方法的分析与研究[J]. 中国安全科学学报, 2004(7): 89-92.
HAN Li, MEI Qiang, LU Yumei, et al. Analysis and study on AHP-fuzzy comprehensive evaluation[J]. China Safety Science Journal, 2004(7): 89-92.
[26] 邓雪, 李家铭, 曾浩健, 等. 层次分析法权重计算方法分析及其应用研究[J]. 数学的实践与认识, 2012, 42(7): 93-100.
DENG Xue, LI Jiaming, ZENG Haojian, et al. Research on computation methods of AHP wight vector and its applications[J]. Mathematics in Practice and Theoty, 2012, 42(7): 93-100.
[27] 陈晓鹏, 侯倩. 服装销售网页色彩效果的模糊综合评价[J]. 丝绸, 2014, 51(7): 42-46.
CHEN Xiaopeng, HOU Qian. Fuzzy comprehensive evaluation of color effect of clothing sales webpage[J]. Journal of Silk, 2014, 51(7): 42-46.
[28] 张辉, 信晓瑜, 马嘉琪, 等. 基于层次分析法的新疆尼雅出土服饰虚拟复原效果评价[J]. 毛纺科技, 2022, 50(4): 66-72.
ZHANG Hui, XIN Xiaoyu, MA Jiaqi, et al. Evaluation of virtual restoration effects of Niyang excavated costumes in Xinjiang based on analytic hierarchy process[J]. Wool Textile Journal, 2022, 50(4): 66-72.
[29] 张英姿, 张技术. 基于层次分析法的服装审美评价维度权重计算[J]. 毛纺科技, 2019, 47(5): 81-85.
ZHANG Yingzi, ZHANG Jishu. Weight calculation of apparel aesthetics evaluation dimension based on AHP[J]. Wool Textile Journal, 2019, 47(5): 81-85.
[1] 程碧莲, 蒋高明, 李炳贤. 三维服装虚拟展示技术的研究进展[J]. 纺织学报, 2024, 45(05): 248-257.
[2] 赖安琪, 蒋高明, 李炳贤. 全成形毛衫花式结构三维仿真[J]. 纺织学报, 2023, 44(02): 103-110.
[3] 周艳红, 江红霞, 陈玲. 基于AutoCAD参数化中式嫁衣样板的自动生成[J]. 纺织学报, 2022, 43(09): 175-181.
[4] 詹必钦, 李玉贤, 董智佳, 丛洪莲. 全成形双层结构针织服装部件化虚拟展示研究[J]. 纺织学报, 2022, 43(08): 147-152.
[5] 邹亚男, 夏风林, 董智佳, 黄梦婷, 储开元. 经编全成形脖套的结构设计与工艺实现[J]. 纺织学报, 2021, 42(12): 76-80.
[6] 江红霞, 黄智威, 刘基宏. 基于模块化划分的旗袍虚拟展示[J]. 纺织学报, 2021, 42(05): 138-142.
[7] 宋莹. 旗袍个性定制和展示系统的交互设计[J]. 纺织学报, 2021, 42(04): 144-148.
[8] 刘海桑, 蒋高明, 董智佳. 基于Web的少梳经编色织物仿真与虚拟展示[J]. 纺织学报, 2021, 42(02): 87-92.
[9] 张卓, 丛洪莲, 蒋高明, 董智佳. 基于交互式遗传算法的Polo衫快速款式推荐系统[J]. 纺织学报, 2021, 42(01): 138-144.
[10] 潘博, 钟跃崎. 基于二维图像的三维服装重建[J]. 纺织学报, 2020, 41(04): 123-128.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!