Defect detection is a crucial direction of deep learning, which is suitable for industrial inspection of product quality in strip steel. As the strip steel production line continuously outputs products, it is necessary to take corresponding measures for the type of defect, once a subtle quality problem is found on steel strips. We propose a new defect area detection and classification method for automation strip steel defect detection. In order to eliminate the way of insufficient data in industrial production line scenarios, we design a transfer learning scheme to support the training of defect region detection. Subsequently, in order to achieve a more accurate classification of defect categories, we designed a deep learning model that integrated the detection results of defect regions and defects feature extraction. After applying our method to the test set and production line, we can achieve extremely high accuracy, reaching 87.11%, while meeting the production speed of the production line compared with other methods. The accuracy and speed of the model realize automatic quality monitoring in the manufacturing process of strip steel.

In the production of the galvanized cold-rolled steel sheets used for stamping car body parts, in-situ and real-time defective detecting is crucial for quality control, in which various types of defects will inevitably occur. It is challenging to improve the accuracy of defect image classification by appropriate means to assist the manual screening process better. Defects under actual production conditions are often not prominent enough in defect characteristics, and there may be a significant similarity between different defect categories. To eliminate this weakness, we propose a data-driven faulty detection model named Steel Faulty Detection Attention Net (SFDANet) that uses images of the galvanized steel surface as input to identify whether the product is qualified and automatic classification of defect types instantaneously. This method can shorten product inspection time and improve production line efficiency automatically. In addition, the attention mechanism is utilized, enhancing the performance of SFDANet. Compared with the baseline that applied the ResNet method, SFDANet achieves a noticeable improvement in the classification accuracy of the test data. The well-trained model can successfully show an improved performance than the baseline models on the multiple types of faulty. Enhanced by SFDANet with high classification accuracy, the defect rate of products is significantly reduced, and the production speed of the production line is significantly improved.

不锈钢冷轧板作为钢铁产业的主要产品形式直接应用于电子、机械、化工、家电、汽车、桥梁、船舶、核电、航空航天等领域.太钢作为全球最大的不锈钢生产基地,通过人工应用表面检测仪检验钢板的作业已经持续了近十年,随着产能的增大,原有的表面质量检验方式已不满足生产产品的评价要求,在表面检测仪应用的基础上开发的表自动判级系统,已成功地在连轧机组投入使用,在国内不锈钢冷轧行业尚属首次.

提出一种基于保相位变换(PHOT)的纹理抑制方法对花纹板表面进行处理,抑制规律纹理,突出非规律纹理即缺陷区域.首先对花纹板图像进行预处理,主要包括降采样、亮度补偿及水痕标记,对预处理后的图像进行FFT变换,幅度均一化及FFT反变换,得到仅保留相位信息的频域图像.对得到频域图像求梯度以放大缺陷和正常区域的差异,并采用马氏距离对梯度图进行二值化操作.最后对二值化图像进行包括缺陷重构和去水痕在内的后处理操作,从而实现花纹板表面缺陷检测.经测试,该算法对花纹板表面存在的花纹残缺、花纹错乱、翘皮和油斑四种缺陷都有较好的检出效果.

A surface inspection system for hot-rolled strip steel was developed independently to meet the urgent needs for detection of surface defects of hot rolled strip steel.This system was designed based on the machine vision technology,using line-scan CCD cameras as surface image acquisition devices.A super bright LED light was developed independently,which can solve the problem of inhomogeneous illumination from a long distance at a high temperature.The high-efficiency image-processing and objects-recognizing algorithms are used in the system so as to guarantee the real-time image processing and precise recognition of the surface defects.The eight monthes on-line operation proves that the system has a higher accuracy for defect detection and classification,together with real time and reliability.

FS-Bagging,a new feature selection based ensemble classifier algorithm,is proposed to cope with common classifier problems elicited by large attribute sets.The new classifier algorithm includes an application of both ReliefF algorithm and Pearson algorithm in an initial step,where a sub-optimal attribute set is derived from the original feature set.Subsequently,the sub-optimal attribute set is subjected to a random selection process,in which a number of feature subsets are grouped according to Bagging algorithm.The feature subsets and their corresponding training data sets are further used to construct a group of classifiers,and the final classification is achieved by simple majority voting results of each individual classifier within this ensemble.The feature subsets and their corresponding training data sets are further used to construct a group of classifiers,and the final classification is achieved by simple majority voting results of each individual classifier within this ensemble.A test on this new classifier with 18 main types of surface defect in strip steel production(data retrieved from a surface inspection system in a Chinese steel company)shows positive improvement over classical Bagging classifiers in terms of efficiency and accuracy in classification.