基于易语言的毛呢织物组织结构识别

Identification of wool fabric structure by the E-language software

为改善生产过程中人工分析织物组织结构的困境,提高企业对毛呢织物自动识别其组织结构的效率,收集了平纹、斜纹、缎纹类毛呢织物,以棉织物图像预处理技术为基础,采用最大类间方差法(Otsu)、灰度共生矩阵以及模糊聚类算法(FCM算法)等对毛呢织物图像组织特征提取、分类,将经处理的毛呢织物图像与组织库中的组织结构进行比对,得出与之相匹配的组织及概率。最后通过编写易语言程序,实现毛呢织物组织结构分析预测。结果表明:采用易语言对三原素色毛呢织物进行分析,试验所得经纬向密度误差低于5%,织物组织类型与实际组织基本一致,即该方法对常见的毛呢织物组织结构的识别具有一定参考性。

In order to improve the dilemma of artificial analysis of fabric structure in production process and improve the efficiency of automatic analysis of the weave structure of woolen fabrics in enterprises,plain weave,twill and satin weave were collected to be investigated.Based on the cotton fabric image pretreatment technology,the structure characteristics of woolen fabric images were extracted and classified by Otsu(maximum variance method),Gray-level co-occurrence matrix and FCM(fuzzy clustering algorithm).The processed woolen fabric image was compared with the structure in the structure database,and the matching structure and probability were obtained.Finally,the analysis and prediction of the woolen fabric structure was realized fluently by E-language.The results show that the error of warp and weft density of conventional fabric is less than 5%,and the obtained weave type is almost consistent with the actual weave.The method can be used as a reference for identifying the weave structure of plain,twill and satin woollen fabrics.