棉纺织技术2025，Vol.53Issue(8)：56-63,8.
畲族服饰色彩提取方法的优化
Optimization of color extraction method for She ethnic costumes
摘要
Abstract
In order to accurately extract the colors of She ethnic costumes,an optimized color extraction method based on multi-variate fuzzy C-means(MFCM)clustering algorithm and K-means++algorithm was proposed.To address the issue of difficulty in extracting high brightness,the bilateral filtering algorithm was first used to smooth the fabric patterns in the image.Then,the HSB color space was selected to distinguish the differences in high brightness colors,accurately describe the visual perception effect of the image.Subsequently,the MFCM clustering algorithm was utilized to determine the optimal number of clusters,achieving accurate segmentation of the image.The K-means++algorithm was used to optimize the selection of initial cluster centers,the class values representing the pattern colors were obtained,and thus the extracted pattern was more accurate.Finally,the algorithm result image of the complete pattern was compared with the original cultural relic image for verification,and quantitative evaluation was conducted using peak signal-noise ratio(PSNR)and structural similarity index(SSIM)to calculate the similarity between the feature color reconstruction image and the original image.The results showed that the PSNR value of the complete pattern was reached up to 33.261 8 dB and the SSIM value was reached 0.976 0,which indicated that the colors of She ethnic traditional costumes could be accurately extracted.The optimized method proposed in this study showed high accuracy and consistency in color extraction.关键词
畲族服饰/多变量模糊C均值聚类算法/K-means++算法/色彩提取/图像处理技术/HSB颜色空间Key words
She ethnic costumes/multi-variable fuzzy C-means clustering algorithm/K-means++algorithm/color extraction/image processing technology/HSB color space分类
轻工纺织引用本文复制引用
卢思宇,胡燕..畲族服饰色彩提取方法的优化[J].棉纺织技术,2025,53(8):56-63,8.基金项目
国家社会科学基金年度项目(22BMZ064) （22BMZ064）
教育部高校学生司供需对接就业育人项目(20230105630) （20230105630）
