基于图像处理法的空气变形纱结构和细度不匀研究

Study on Structure and Unevenness of Air Textured Yarn by Image Processing Method

为定量表征空气变形纱的结构特征,采用图像处理的方法,建立了空气变形纱纱芯-表层几何结构模型,并提出用纱线圈弧率、圈弧指数及表观不匀率3项指标来定量表征其结构特征和细度不匀.分析结果表明:根据纱线二值化图像逐行扫描像素分布曲线,可较好地提取纱线纱芯部分;采用图像法获得的纱线表观不匀率比传统纱线条干均匀度仪测得的条干不匀率值要大,表明空气变形纱的外观形态不匀率大于其质量不匀率;与常规毛羽仪测试结果相比,图像法不会改变纱线表面圈弧形态,并可减少对纱线表面细小圈弧的丢失,能较好地反映空气变形纱表面圈弧分布规律.

In order to study the structure of air textured yarn quantitatively, the core-and-effect structure model of air textured yarn was built based on image processing method. Three indexes including loop ratio, loop index and yarn apparent unevenness, were proposed to characterize the structure and evenness of air textured yarn quantitatively. The results show that the core structure of yarn can be obtained effectively from the binary image of yarn by scanning in lines. Unevenness calculated by image method is higher than that obtained by traditional yarn unevenness test, indicating a greater yarn morphological unevenness than yarn mass unevenness for air textured yarn. Compared to traditional yarn hairiness test, the loop distribution of air textured yarn can be characterized effectively by image processing method by keeping the original shape of loops and reducing the loss of fine loops statistically.