基于数字图像处理的醋酸纤维丝束在线厚度检测

Online Thickness Detection of Acetate Fiber Bundles Based on Digital Image Processing

针对醋酸纤维丝束厚度检测缺乏成熟的非接触式检测方案的现状,基于丝束厚度与本身均匀性以及所受拉力有关,不同厚度的丝束区域透光性不同,呈现在图像中的灰度不同的原理,提出基于数字图像处理的丝束在线厚度检测方法。首先采用图像滤波去除噪声;接着通过Otsu阈值分割实现目标和背景的精确分离;随后进行连通区域划分,不同区域分别提取灰度信息,并结合光学领域知识,构建在线厚度检测模型,实现了对丝束厚度的准确检测。实验结果表明:该方法实现了对丝束区域的划分和对丝束厚度的计算,其厚度计算值与实际测量值的偏差不超过8.00%,满足实际检测需求。

Considering the fact that lack of mature non-contact detection scheme for measuring the thickness of acetate fiber bundles which related to its uniformity and tensile force,both light transmittance of the acetate fiber bundle with various thickness and gray level in the image is different,an online thickness detection method based on digital image processing was proposed.in which,having image filtering used to remove noise and then,having Otsu threshold segmentation adopted to accurately separate both object and background and as well as having connected regions divided;in addition,having gray information extracted from different regions;and combined with the optical domain knowledge,having online thickness detection model constructed to realize accurate detection of the thickness of the filament bundle.The experimental results show that,this method can realizes division of the filament bundle area and calculation of the thickness of the filament bundle;and the deviation between calculated thickness value and actual measured thickness value is less than 8.00%,which meets actual detection requirements.