Abstract
Steel defect detection is used to detect defects on the surface of the steel and to improve the quality of the steel surface. However, traditional image detection algorithms cannot meet the detection requirements because of small defect features and low contrast between background and features about steel surface defect datasets. A novel recognition algorithm for steel surface defects based on improved deep learning network models using feature visualization and quality evaluation is proposed in this paper. Firstly, the VGG19 is used to pre-train the steel surface defect classification task and the corresponding DVGG19 is established to extract the feature images in different layers from defects weight model. Secondly, the SSIM and decision tree are used to evaluate the feature image quality and adjust the parameters and structure of VGG19. On this basis, a new VSD network is obtained and used for the classification of steel surface defects. Comparing with ResNet and VGG19 methods, experiment results show that the proposed method markedly can improve the average accuracy of classification, and the model is able to converge quickly, which can be good for steel surface defect recognition using VSD network model of feature visualization and quality evaluation.
Highlights
Steel quality control is a hard problem in quality management
To improve the classification accuracy of industrial hot-rolled steel surface defects, this paper introduces an algorithm that visualizes the underlying features of a defect and adjusts the network based on the quality evaluation results of the featured image
Due to the fewer amounts of pixels in the picture of the industrial steel defect dataset and the lower contrast between features and background, it is difficult for traditional deep learning methods to have excellent cores on the steel defect dataset
Summary
Steel quality control is a hard problem in quality management. The quality of the steel affects the cost of the product and has a great impact on the subsequent processing accuracy. For different detection objects and detection background, the deep learning algorithms have sufficient flexibility to achieve the detection target [14]–[20]. In 1990, they optimized the framework and proposed a new structure called LeNet-5 [25], which consists of a multi-layer neural network and shows excellent performance in handwritten digital classification. S. Guan et al.: Steel Surface Defect Recognition Algorithm Based on Improved Deep Learning Network Model similar but deeper structure than LeNet-5, and the classification accuracy was significantly improved compared with the previous one. To improve the problem caused by the linear structure [28], a nonlinear deep residual learning framework was introduced in ResNet and performed well in image classification tasks. The current convolutional neural network models for the image classification task in the steel background are too complicated and redundant
Sign-in/Register to view highlights & summary 
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call