基于卷积神经网络的磁瓦缺陷检测研究
|
摘要
针对磁瓦缺陷种类多样性及无法准确描述其缺陷的问题,提出一种基于卷积神经网络的缺陷检测方法。构建缺陷类型的数据集,并对数据集中的图像进行预处理;设置卷积神经网络模型参数,训练缺陷分类器;通过训练结果完成对缺陷图像的识别并标注缺陷类型。实验结果表明,该方法检测的准确性和实时性均优于传统检测方法,具有非常好的鲁棒性,为工业生产的实际应用提供了可靠的依据。
In view of the diversity and inaccurate description of magnetic tile defects, a defect detection method based on the convolutional neural network is proposed. Firstly, the data set of defect types is constructed, and the images in the data set are preprocessed. Secondly, the convolutional neural network model parameters are set to train the defect classifier. Finally, the defect images are recognized and the defect types are labeled by the training results. The experimental results show that the accuracy and real-time performance of this method are better than traditional detection methods. It also has good robustness, which provides a reliable basis for the practical application of industrial production.
In view of the diversity and inaccurate description of magnetic tile defects, a defect detection method based on the convolutional neural network is proposed. Firstly, the data set of defect types is constructed, and the images in the data set are preprocessed. Secondly, the convolutional neural network model parameters are set to train the defect classifier. Finally, the defect images are recognized and the defect types are labeled by the training results. The experimental results show that the accuracy and real-time performance of this method are better than traditional detection methods. It also has good robustness, which provides a reliable basis for the practical application of industrial production.
引文
[1] YANG C L,LIU P Y,YIN G F,et al.Crack detection in magnetic tile images using nonsubsampled shearlet transform and envelope gray level gradient[J].Optics & Laser Technology,2017,90:7-17.
[2] LEI J,GAO X,FENG Z L,et al.Scale insensitive and focus driven mobile screen defect detection in industry[J].Neurocomputing,2018,294:72-81.
[3] SAEED H H,AHMAD A,FARSHAD B.Automatic detection and classification of the ceramic tiles’ surface defects[J].Pattern Recognition,2017:174-189
[4] HUANG Y B,QIU C Y.Surface defect saliency of magnetic tile[C].Proceedings of IEEE International Conference on Automation Science and Engineering (CASE),2018:1-12.
[5] 黄兴.磁瓦缺陷自动检测关键技术研究[D].南京:东南大学,2017.
[6] 胡浩,李俊峰,沈军民.基于机器视觉的小磁瓦表面微缺陷检测方法研究[J].机电工程,2019,36(2):117-123.
[7] LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition[C].Proceedings of the IEEE,1998,86(11):2278-2324.
[8] LECUN B B,DENKER J S,HENDERSON D,et al.Handwritten digit recognition with a back-propagation network[C].Advances in neural information processing systems.Citeseer,1990:396-404.
[9]于科为.基于卷积神经网络的工件缺陷检测研究[J].信息与电脑(理论版),2018(21):7-9.
[2] LEI J,GAO X,FENG Z L,et al.Scale insensitive and focus driven mobile screen defect detection in industry[J].Neurocomputing,2018,294:72-81.
[3] SAEED H H,AHMAD A,FARSHAD B.Automatic detection and classification of the ceramic tiles’ surface defects[J].Pattern Recognition,2017:174-189
[4] HUANG Y B,QIU C Y.Surface defect saliency of magnetic tile[C].Proceedings of IEEE International Conference on Automation Science and Engineering (CASE),2018:1-12.
[5] 黄兴.磁瓦缺陷自动检测关键技术研究[D].南京:东南大学,2017.
[6] 胡浩,李俊峰,沈军民.基于机器视觉的小磁瓦表面微缺陷检测方法研究[J].机电工程,2019,36(2):117-123.
[7] LECUN Y,BOTTOU L,BENGIO Y,et al.Gradient-based learning applied to document recognition[C].Proceedings of the IEEE,1998,86(11):2278-2324.
[8] LECUN B B,DENKER J S,HENDERSON D,et al.Handwritten digit recognition with a back-propagation network[C].Advances in neural information processing systems.Citeseer,1990:396-404.
[9]于科为.基于卷积神经网络的工件缺陷检测研究[J].信息与电脑(理论版),2018(21):7-9.