基于K-Means聚类算法的秀场服装用色风格  

Color Style of Fashion Show Clothing Based on K-Means Clustering Algorithm

作　　者：陈郁 雷聪聪 CHEN Yu;LEI Congcong(School of Textiles and Fashion,Shanghai University of Engineering Science,Shanghai 201620,China)

机构地区：[1]上海工程技术大学纺织服装学院,上海201620

出　　处：《服装学报》2024年第3期268-273,共6页Journal of Clothing Research

基　　金：上海市教育委员会东方学者项目(TP2017074)。

摘　　要：针对传统手抄方法进行秀场色彩分析效率低下的问题,使用自动抠像算法提取秀场中的服装,根据服装图片RGB属性值,采用K-means聚类算法实现服装色彩聚类,得出当季服装用色风格。结果表明,秀场服装色彩自动聚类的方法较人工统计法效率大幅度提高,输出的分析结果可以更直观呈现当季知名品牌用色风格和趋势,为设计师确定服装用色提供参考。In response to the low efficiency of traditional hand-copied fashion show color analysis,this article used an auto-matic image extraction algorithm to extract clothing from the fashion show.Based on the RGB attribute values of clothing images,K-means clustering algorithm was used to cluster clothing colors and obtain the color style for the current season's clothing.The results show that the automatic color clustering method for fashion shows exhibit a significantly higher efficiency compared to manual statistical methods,and the output analysis results can intuitively present the color styles and trends of well-known brands in the current season,and can provide a reference for designers to determine the color of clothing.

关 键 词：服装色彩 色彩统计 聚类算法 秀场服装 用色风格 

分 类 号：F768.3[经济管理—产业经济]