激光与光电子学进展, 2020, 57 (20): 201507, 网络出版: 2020-10-14   

基于机器视觉的编织袋缺陷在线检测方法 下载: 868次

Online Detection Method of Woven Bag Defects Based on Machine Vision
迟欢 *
作者单位
北京科技大学天津学院机械工程系, 天津 301830
摘要
针对人工检测编织袋缺陷的正确率低与效率较低的问题,提出一种高效的在线检测编织袋缺陷方法。该方法在线采集编织袋图像并进行图像处理,消除干扰项,准确检测编织袋的缺陷。使用均值滤波器、灰度开闭操作对图像进行预处理,消除图像中干扰缺陷检测的黑白条纹与灰度不均匀,降低噪声。使用差分图像二值化对图像进行背景分割,提取出孔洞缺陷、拉丝缺陷,以及过大的丝线缝隙、褶皱和黑色物。同时,进行开闭运算处理,将断裂的缺陷连接起来并消除过大的丝线缝隙,避免小缺陷的漏检。利用特征提取与缺陷检测消除褶皱和黑色物的干扰,检测出孔洞和拉丝缺陷。实验结果表明,500个试样检测的平均正确检测率达到97.20%,检测效率为720 m/h,检测结果正确率高,效率高。
Abstract
To solve the problem of low accuracy and low efficiency in manual detection of woven bag defects, an efficient online detection method for woven bag defects is proposed. The method collects images of woven bags online and performs image processing to eliminate interference items and accurately detect defects in woven bags. The image is preprocessed by using the mean filter and gray-scale open and close operations to eliminate black and white stripes and gray-scale unevenness that interfere with defect detection in the image, and reduce noise. Use differential image binarization to perform background segmentation on the image, and extract hole defects, wire drawing defects, and excessive wire gaps, wrinkles, and black objects. At the same time, open and close operation is used to connect the broken defects and eliminate the excessive wire gaps in the silk thread, so as to avoid the omission of small defects. Feature extraction and defect detection are used to eliminate the interference of folds and black objects, and detect holes and drawing defects. Experimental results show that the average correct detection rate of 500 samples reaches 97.20%, the detection efficiency is 720 m/h, and the detection accuracy and efficiency are high.

迟欢. 基于机器视觉的编织袋缺陷在线检测方法[J]. 激光与光电子学进展, 2020, 57(20): 201507. Huan Chi. Online Detection Method of Woven Bag Defects Based on Machine Vision[J]. Laser & Optoelectronics Progress, 2020, 57(20): 201507.

本文已被 3 篇论文引用
被引统计数据来源于中国光学期刊网
引用该论文: TXT   |   EndNote

相关论文

加载中...

关于本站 Cookie 的使用提示

中国光学期刊网使用基于 cookie 的技术来更好地为您提供各项服务,点击此处了解我们的隐私策略。 如您需继续使用本网站,请您授权我们使用本地 cookie 来保存部分信息。
全站搜索
您最值得信赖的光电行业旗舰网络服务平台!