基于FCM算法的有色涤纶纤维混合比测定及影响因素分析

Measurement of colored polyester fiber mixing ratio and analysis of influencing factors based on FCM algorithm

为解决色纺纱生产中测配色存在的准确性不足和耗时问题,基于模糊C均值(fuzzy Cmeans,FCM)聚类算法测定有色涤纶纤维混色样品中各有色纤维的比例,以提高配色效率。用扫描仪采集样品图像,再利用FCM算法在LAB颜色空间中对图像进行聚类分析,通过统计聚类分析结果中各彩色像素的百分比间接得到配色比例,并讨论聚类中心数、扫描背景、扫描分辨率、扫描面积和纤维混合次数对聚类分析结果的影响。结果表明:就现有试验条件而言,在白色背景下,用2 400dpi的分辨率采集30mm×30mm的样品图像,用聚类中心数目为7的FCM算法分析时,聚类分析的效率较快且结果较稳定和准确。由此可见,通过图像处理技术分析涤纶色纺纱的配色比例,可减少配色时间,提高配色质量。

To solve the lack of accuracy and time-consuming problem of color matching in color spinning production,the fuzzy C-means(FCM)clustering algorithm was used to determine the proportion of each colored fiber in the mixed sample to improve the color matching efficiency.The sample images were collected with a scanner and clustered in LAB color space using the FCM algorithm,and the color matching ratio was obtained indirectly by counting the percentage of colored pixels in the clustering results.The results show that,under the existing test conditions,the clustering analysis is faster and more stable and accurate when the FCM algorithm with the number of clustering centers of 7 is used to analyze the 30 mm×30 mm sample images with 2400 dpi resolution on a white background.It shows that analyzing the color matching ratio of polyester color-spun yarns by image processing technique can reduce the color matching time and improve the color matching quality.