基于图像处理的织物覆盖系数检测  

Inspection on the cover factors of fabric based on image processing

作　　者：陈阳[1] 陈霞[1,2] 汪军[1] 

机构地区：[1]东华大学纺织学院,上海201620 [2]东华大学产业用纺织品教育部工程研究中心,上海201620

出　　处：《西安工程大学学报》2016年第1期16-20,共5页Journal of Xi’an Polytechnic University

基　　金：中央高校基本科研业务费专项资金资助项目(14D110130)

摘　　要：针对短纤纱织物图像中毛羽的干扰和图像背景不匀,将Canny边缘分析和局部阈值分割法结合用于检测织物的覆盖系数.首先根据纱线中心线将织物图像划分成许多局部图像,利用Canny算子检测局部图像中孔隙的边缘,确定有效局部阈值并转换成二值图像.然后将局部二值图像形态学处理后融合得到完整的织物二值图像,计算覆盖系数.最后对3种自织的平纹棉织物的覆盖系数进行检测,并与理论计算的覆盖系数比较,两者相关系数达0.99,说明此方法可以较好地检测短纤纱织物的覆盖系数.To reduce the influence of hairiness interference and background unevenness on sta- ple fabric image, the local threshold segmentation and Canny edge analysis were applied to detect the cover factors of fabric. Firstly,the fabric image was divided into a series of local images by yarn centerlines. Using Canny edge detection to obtain effective local threshold, these local images were transformed into binary images. After morphological processing, these local bina ry images were merged into a complete binary image and cover factor was calculated. Finally, the actual cover factor of three plain fabric samples was compared with theoretical values. The correlation coefficient between the actual and theoretical values was 0.99. The result shows that the method in this paper can detect the cover factors of staple yarn fabric effectively.

关 键 词：短纤纱 机织物 覆盖系数 图像处理 CANNY边缘检测 局部阈值 

分 类 号：TS101.9[轻工技术与工程—纺织工程]