Scratches are one of the most common defects in industrial manufacturing. Weak scratches in the industrial environment have an ambiguous edge, low contrast, large span, and unfixed shape, which bring difficulty for automatic defect detection. Recently, many existing visual inspection methods based on deep learning cannot completely and effectively inspect industrial weak scratches due to the lack of discriminative features and sufficient spatial detail. In this article, a novel DeepScratchNet is proposed for automatic weak scratch detection by aggregating rich multidimensional feature for scratch representation. To obtain rich features, a pretrained ResNet block as a feature extractor is proposed in this article. To highlight features of scratch and weaken the noise, an attention feature fusion block (AFB) is proposed, which densely fuses high-level semantic features with low-level detail features using dual-attention mechanism. Due to the long span and connectivity of the weak scratches, a context fusion block (CFB) is proposed to learn the complete context. To further improve the scratch segmentation performance, the auxiliary loss is integrated into the proposed network. The proposed DeepScratchNet outperforms the traditional and other state-of-the-art deep learning-based methods on three given real-world industrial data sets with mIoU over 0.8005, 0.812, and 0.9286. The experimental results demonstrate that DeepScratchNet achieves good generalization capabilities.

Traditional image steganography techniques complete the steganography process by embedding secret information into cover images, but steganalysis tools can easily detect detectable pixel changes that lead to the leakage of confidential information. The use of a generative adversarial network (GAN) makes it possible to embed information using a combination of information and noise in generating images to achieve steganography. However, this approach is usually accompanied by issues such as poor image quality and low steganography capacity. To address these challenges, we propose a steganography model based on a novel information-driven generative adversarial network (IDGAN), which fuses a GAN, attention mechanisms, and image interpolation techniques. We introduced an attention mechanism on top of the original GAN model to improve image accuracy. In the generation model, we replaced some transposed convolution operations with image interpolation for better quality of dense images. In contrast to traditional steganographic methods, the IDGAN generates images containing confidential information without using cover images and utilizes GANs for information embedding, thus having better anti-detection capability. Moreover, the IDGAN uses an attention mechanism to improve the image details and clarity and optimizes the steganography effect through an image interpolation algorithm. Experimental results demonstrate that the IDGAN achieves an accuracy of 99.4%, 95.4%, 93.2%, and 100% on the MNIST, Intel Image Classification, Flowers, and Face datasets, respectively, with an embedding rate of 0.17 bpp. The model effectively protects confidential information while maintaining high image quality.

Scratches as the major defects in precision optical components are caused inevitably in the manufacturing process, which are harmful to the whole optical system. Most scratches on the surface of optical components are weak scratches with low contrast and uneven distribution of grayscale, which poses a significant problem for inspection. In this paper, an end-to-end weak scratch inspection method based on novel scratch enhancement methods and convolutional neural network (CNN) is proposed for optical components. To enhance weak scratches, a local maximum index (LMI) module and a direction sensitive convolution (DSC) module are proposed to generate multilevel-feature maps using prior knowledge about scratch. Different from previous works utilizing the raw dark-field image as network input, these multilevel features are used as the inputs of encoder-decoder module for training. After training, the whole inspection model can infer weak scratches from raw dark-field test images in an end-to-end manner. Experimental results show that the proposed model achieves pixel accuracy of 92.48% and IoU at 77.27% on the test dataset. It outperforms the networks without adding prior knowledge, which shows that prior knowledge is much helpful for weak scratch inspection. Moreover, compared with other classical methods and CNN-based methods, the proposed method achieves the best performance in the weak scratch inspection.

In high-power laser instruments, optical elements play a significant role. Particles on the optical element degrade the system performance and even cause damage to the optical element. In this article, a particle inspection model based on self-supervised convolutional neural networks (CNNs) and transfer learning is proposed. The self-supervised network that is built on a rotation-flip-invariant pretext task is used to transform the image from grayscale feature to rotation-flip-invariant feature. Then, the learned feature is transferred to the central-pixel classification network that is fine-tuned on a small labeled dataset. The experiments show that the classification accuracy of our proposed method is 97.90%, which is higher than the other compared methods. For the whole image prediction, through feature reuse and pointwise convolution, the central-pixel classification network is adapted to the particle inspection network efficiently with minor changes. Since the method utilizes massive unlabeled data and is fine-tuned on a small number of labeled samples, it has the potential to be used in industrial production.

With the application and development of Internet technology, network traffic is growing rapidly, and the situation of network security is becoming more and more serious. As an important way to protect network security, abnormal traffic detection has been paid more and more attention. In this paper, the uncertainty of the samples in the abnormal traffic detection dataset is studied. Combining the three-way decision idea with the random forest algorithm, a three-way selection random forest optimization model for abnormal traffic detection is proposed. Firstly, the three-way decision idea is integrated into the random selection process of feature attributes, and the attribute importance based on decision boundary entropy is calculated. The feature attributes are divided into the normal domain, abnormal domain, and uncertain domain, and the three-way attribute random selection rules are designed to randomly select the feature attributes that conform to the rules from different domains. Secondly, the classifier evaluation function is constructed by combining pure accuracy and diversity, and the anomaly traffic detection base classifier with a high evaluation value is selected for integration to eliminate the unstable factors caused by randomness in the process of base classifier generation. Thirdly, the optimal node weight combination of the base classifier is obtained by iterative calculation of the gray wolf optimization algorithm to further improve the prediction effect and robustness of the model. Finally, the model is applied to the abnormal traffic detection dataset. The experimental results show that the prediction accuracy of the three-way selection random forest optimization model on CIC-IDS2017, KDDCUP99, and NSLKDD datasets is 96.1%, 95.2%, and 95.3%, respectively, which has a better detection effect than other machine learning algorithms.