抓毛织物表面质量的视觉检测方法

Visual Inspection Method for Evaluating Surface Quality of Fluff Fabric

针对抓毛织物表面质量的量化检测问题,提出一种基于计算机视觉的抓毛织物表面质量检测方法。采用切向背光法获取抓毛织物的轮廓图像,根据绒毛区域的灰度直方图特征,对绒毛区域进行分割,再利用形态学算法对绒毛区域内出现的空洞进行填充,从而得到完整的绒毛区域。通过Freeman编码提取绒毛区域的两侧边缘坐标,利用最小二乘算法拟合绒毛下边缘得到绒毛厚度基准线并建立绒毛厚度模型。同时采用统计直方图构建绒毛分布状态参数模型。结果表明:该方法对织物的颜色、底布的纹理不敏感,可以实现对抓毛织物表面质量进行综合量化评定。

Aiming at the problem of quantitative detection of surface quality of fluff fabric,a method based on computer vision for detecting the surface quality of fluff fabric was proposed.The tangential backlight method was used to obtain the contour image of the fluff fabric.According to the gray histogram feature of the fluff region,the fluff region was segmented,and then the morphological algorithm was used to fill the voids in the fluff region.Thus,a complete Fluff area was obtained.The edge coordinates of the two sides of the fluff region were extracted by Freeman coding,and the lower edge of the fluff was fitted by the least squares algorithm to obtain the fluff thickness reference line.As well,the fluff thickness model was established.At the same time,the statistical histogram was adopted to construct the parameter model of fluff distribution state.The experimental results show that the method is not sensitive to the color of the fabric and the texture of the base fabric,and the surface quality of the fluff fabric can be comprehensively evaluated by this method.