The surface defect detection of industrial products has become a crucial link in industrial manufacturing. It has a series of chain effects on the control of product quality, the safety of the subsequent use of products, the reputation of products, and production efficiency. However, in actual production, it is often difficult to collect defect image samples. Without a sufficient number of defect image samples, training defect detection models is difficult to achieve. In this paper, a defect image generation method DG-GAN is proposed for defect detection. Based on the idea of the progressive generative adversarial, $D2$ adversarial loss function, cyclic consistency loss function, a data augmentation module, and a self-attention mechanism are introduced to improve the training stability and generative ability of the network. The DG-GAN method can generate high-quality and high-diversity surface defect images. The surface defect image generated by the model can be used to train the defect detection model and improve the convergence stability and detection accuracy of the defect detection model. Validation was performed on two data sets. Compared to the previous methods, the FID score of the generated defect images was significantly reduced (mean reductions of 16.17 and 20.06, respectively). The YOLOX detection accuracy was significantly improved with the increase in generated defect images (the highest increases were 6.1% and 20.4%, respectively). Experimental results showed that the DG-GAN model is effective in surface defect detection tasks.

Currently, the development of industry is becoming increasingly rapid. Technicalization, informatization and industrialization give the fundamental impetus for industrial development and progress. Nevertheless, there are numerous problems that are hindering industrial progress and threatening human security in the industrial field. The surface defects of the workpieces are one of the primary problems. Moreover, defects of multi-type, mixed and unapparent characteristics presented by workpieces make the detection and classification of workpiece more difficult. Deep convolutional neural networks (DCNN) show strong ability of feature extraction and mines deeper essential features of data because of its features of unique receptive field structure and weights of shared. It can represent original data information well and obtain results more accurately than the traditional methods. But there also remains a problem that conventional DCNN has a huge number of parameters and computation, which brings great pressure to the equipment in terms of computing power, memory, speed and so on. Based on this situation, the optimization methods of CNNs model in the aspects of data, structure, algorithm are summarized. Related lightweight structures and networks are also summarized in this paper. The purpose of these work is to reduce the number of parameters and computation and improve the training performance. At the same time, the research on defect classification of workpieces based on traditional machine learning and deep learning model is reviewed, and the research on defect classification of workpieces based on deeply optimized CNNs is referred and prospected.

In the field of industry, defect detection based on YOLO models is widely used. In real detection, the method of defect detection of insulative baffles is artificial detection. The work efficiency of this method, however, is low because the detection is depends absolutely on human eyes. Considering the excellent performance of YOLOx, an intelligent detection method based on YOLOx is proposed. First, we selected a CIOU loss function instead of an IOU loss function by analyzing the defect characteristics of insulative baffles. In addition, considering the limitation of model resources in application scenarios, the lightweight YOLOx model is proposed. We replaced YOLOx’s backbone with lightweight backbones (MobileNetV3 and GhostNet), and used Depthwise separable convolution instead of conventional convolution. This operation reduces the number of network parameters by about 42% compared with the original YOLOx network. However, the mAP of it is decreased by about 0.8% compared with the original YOLOx model. Finally, the attention mechanism is introduced into the feature fusion module to solve this problem, and we called the lightweight YOLOx with an attention module LA_YOLOx. The final value of mAP of LA_YOLOx reaches 95.60%, while the original YOLOx model is 95.31%, which proves the effectiveness of the LA_YOLOx model.