基于Basler工业相机图像的丝绸表面缺陷检测

在缺陷检测过程中，由于光照角度和亮度等因素导致采集的图像亮度较小，对比度不明显，使用一般相机采集某些材料表面图像时不能取得较好的效果。工业相机成像质量通常高于普通相机，因此利用Basler工业相机获取丝绸材料表面缺陷，基于图像处理缺陷检测技术提取丝绸表面缺陷图像。首先对图像进行灰度转换，得到单通道灰度图像，以便进行图像二值化处理。由于二值化图像有噪声点，用简单的模糊处理无法有效消除，需对图像进行开运算。将图像腐蚀后再进行膨胀，基本上可解决图像噪声问题，根据最后提取的图像轮廓进行筛选，提取出轮廓缺陷。经过检测发现，该算法对工业工件表面缺陷提取更准确，效果更好，可高效检测丝绸表面缺陷，提高生产效率。

Silk Surface Defect Detection Based on Basler Industrial Camera Image

In the process of defect detection， due to some reasons such as illumination angle and brightness， the brightness of the captured images is relatively small， and the contrast is not obvious. Cameras with general performances can not achieve good results for some materials' surface image acquisition. Industrial cameras typically have higher imaging quality than normal cameras. This paper mainly uses Basler industrial camera to obtain the defects of silk material surface， and extracts the image of silk surface defect obtained based on image processing defect detection technology. The first step is to perform grayscale conversion on the image to obtain a single-channel grayscale image， which is beneficial to image binarization. Since the binarized image still has some noise points， the mere use of the blurring of these noise points does not produce good effects. Therefore， the image is opened first， and the image is etched to obtain an image that is eroded to basically solve the problem of image noise， the resulting image contour extraction and screening finally extracts contour defects. After the final test， the algorithm has achieved good results in the extraction of surface defects of some industrial workpieces， and it is more accurate and efficient to detect defects on the silk surface and improve production efficiency.