Deep learning–based fabric defect detection methods have been widely investigated to improve production efficiency and product quality. Although deep learning–based methods have proved to be powerful ...tools for classification and segmentation, some key issues remain to be addressed when applied to real applications. Firstly, the actual fabric production conditions of factories necessitate higher real-time performance of methods. Moreover, fabric defects as abnormal samples are very rare compared with normal samples, which results in data imbalance. It makes model training based on deep learning challenging. To solve these problems, an extremely efficient convolutional neural network, Mobile-Unet, is proposed to achieve the end-to-end defect segmentation. The median frequency balancing loss function is used to overcome the challenge of sample imbalance. Additionally, Mobile-Unet introduces depth-wise separable convolution, which dramatically reduces the complexity cost and model size of the network. It comprises two parts: encoder and decoder. The MobileNetV2 feature extractor is used as the encoder, and then five deconvolution layers are added as the decoder. Finally, the softmax layer is used to generate the segmentation mask. The performance of the proposed model has been evaluated by public fabric datasets and self-built fabric datasets. In comparison with other methods, the experimental results demonstrate that segmentation accuracy and detection speed in the proposed method achieve state-of-the-art performance.
The current state-of-the-art anomaly detection methods based on knowledge distillation (KD) typically depend on smaller student networks or reverse distillation to address vanishing representations ...discrepancy on anomalies. These methods often struggle to achieve precise detection when dealing with complex texture backgrounds containing anomalies due to the similarity between anomalous and non-anomalous regions. Therefore, we propose a new paradigm-Cosine Similarity Knowledge Distillation (CSKD), for surface anomaly detection and localization. We focus on the superior performance of the same deeper teacher and student encoders by the distillation loss in traditional knowledge distillation-based methods. Essentially, we introduce the Attention One-Class Embedding (AOCE) in the student network to enhance learning capabilities and reduce the effect of the teacher-student (T-S) model on response similarity in anomalous regions. Furthermore, we find the optimal models by different classes' hard-coded epochs, and an adaptive optimal model selection method is designed. Extensive experiments on the MVTec dataset with 99.2% image-level AUROC and 98.2%/94.7% pixel-level AUROC/PRO demonstrate that our method outperforms existing unsupervised anomaly detection algorithms. Additional experiments on DAGM dataset, and one-class anomaly detection benchmarks further show the superiority of the proposed method.
Surface inspection is a necessary process of fabric quality control. However, it remains a challenging task owing to diverse types of defects, various patterns of fabric texture, and application ...requirements for detection speed. In this paper, a lightweight deep learning model is therefore proposed to complete the segmentation of fabric defects. The input of the model is a fabric image, and the output is a binary image. Generally known, a deep learning model usually needs much data to update the parameters. Still, as an abnormal phenomenon, fabric defects are unpredictable, which makes it impossible to collect a large number of data. Distinct from other models, the proposed method is a supervised network but does not need manually labeled samples for training. A fake sample generator is designed to simulate the defect image, which only needs the defect-free fabric image. The proposed model is trained with fake samples and verified with real samples. The experimental results show that the model trained with false data is useful and achieves high segmentation accuracy on real fabric samples. Besides, a loss function is proposed to deal with the problem of imbalance between the number of background pixels and the number of defective pixels in the fabric image. Comprehensive experiments were performed on representative fabric samples to verify the segmentation accuracy and detection speed of this method.
Edge detection is a crucial task for computer vision. In this paper, we propose to use both the multi-directional first-order anisotropic Gaussian derivative and the second-order anisotropic Gaussian ...derivative to extract image gray information. The first-order derivative is utilized to determine the gradient direction, while the second-order derivative is used to identify the gradient magnitude. By double filtering of the feature information, the operator’s robustness is improved, and the edge stretching is reduced. The multi-directional filters can obtain enough gradient information to avoid edge missing. Moreover, we propose to use the adaptive thresholds to improve the operator’s generalizability. The aggregate receiver operating characteristic curve shows that the proposed method improves the accuracy of edge detection and exhibits strong robustness.
This study aims to verify whether the inhibitory effect of Sirtuin 3 (SIRT3) on the formation of renal calcium oxalate crystals was mediated through promoting macrophages (Mϕs) polarization. ...Identification and quantification of M1 and M2 monocytes were performed using fluorescence‐activated cell sorting analysis. SIRT3 protein level and forkhead box O1 (FOXO1) acetylation level were measured using western blot analysis. Cell apoptosis of HK‐2 was detected by flow cytometry. Mouse kidney tissues were subjected to Von Kossa staining, terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) staining, and immunohistochemical staining for detection of kidney crystals deposition, apoptosis, and expression of crystal‐related molecules, respectively. The results showed that human peripheral blood monocytes from patients with kidney stone (KS) exhibited decreased M2 monocytes percentage and SIRT3 expression, whereas increased FOXO1 acetylation compared with the normal controls. In vitro assay revealed that SIRT3 overexpression in bone marrow‐derived M0/M1/M2 Mϕs induced M2 polarization and decreased FOXO1 acetylation. Furthermore, FOXO1 knockdown reversed SIRT3‐mediated induction of M2 polarization and inhibition of HK‐2 (human proximal tubular cell line) apoptosis. Further in vivo experiments demonstrated that SIRT3‐overexpressing Mϕs transfusion not only induced M2 polarization, but also alleviated inflammation, apoptosis, and crystals deposition in glyoxylate‐induced KS mice. In conclusion, SIRT3 suppresses formation of renal calcium oxalate crystals through promoting M2 polarization via deacetylating FOXO1.
Our findings support the notion that Sirtuin 3 (SIRT3) suppresses renal inflammation, renal apoptosis, and formation of renal calcium oxalate crystals. The protective role of SIRT3 in kidney stone (KS) disease was mediated, at least in part, through promoting the polarization of macrophages (Mϕs) toward the M2 phenotype via deacetylating forkhead box O1 (FOXO1). This study supports the therapeutic possibility of targeted Mϕs‐phenotype shifting from M1 to M2 in KS disease.
Neural networks have been widely used in color space conversion in the digital printing process. The shallow neural network easily obtains the local optimal solution when establishing ...multi-dimensional nonlinear mapping. In this paper, an improved high-precision deep belief network (DBN) algorithm is proposed to achieve the color space conversion from CMYK to L*a*b*. First, the PANTONE TCX color card is used as sample data, in which the CMYK value of the color block is used as input and the L*a*b* value is used as output; then, the conversion model from CMYK to L*a*b* color space is established by using DBN. To obtain better weight and threshold, DBN is optimized by a particle swarm optimization algorithm. Experimental results show that the proposed method has the highest conversion accuracy compared with Back Propagation Neural Network, Generalized Regression Neural Network, and traditional DBN color space conversion methods. It can also adapt to the actual production demand of color management in digital printing.
Real-time detection of fabric defects is a fairly critical part of industrial production. However, there are still some key issues to be solved in practical detection production, such as low ...detection speed and delays in traditional cloud detection. To address these issues, in this paper, a new detection network architecture, called YOLOV4-TinyS, is proposed. Firstly, the k-medoids clustering algorithm is used to improve the matching of anchor points and ground truths for datasets with great differences. Secondly, the residual structure is changed to reduce the complexity of the network structure, and a depth-separable convolution is used instead of partial convolution to improve the detection speed. Thirdly, the output feature layer is designed with shallow feature fusion to improve the location information extraction capability and use spatial attention and channel attention to improve the network efficiency. Finally, the whole network is trained and tested on four different datasets and extensive experiments show that the network has higher detection accuracy and faster detection speed compared to existing methods. Compared to the original network, YOLOV4-Tiny, the model complexity is reduced by 67.86% and the highest detection accuracy of 99.91% is achieved. Furthermore, the establishment of an efficient fabric inspection system and the validation of the method allows for the fast detection of fabric defects on conveyor belts. Thus, the proposed method has the potential to lay the foundation for the real-time detection of fabric defects and their application in industry.
During spinning, knitting and weaving processes, the yarn apparent evenness is an important factor, which determines the quality of subsequent spinning production and fabric performance. This paper ...presents a powerful method which is based on L0 norm smoothing and the expectation maximization method to detect the yarn apparent evenness. The L0 norm smoothing method is first applied to remove the noise and enhance the yarn apparent evenness diameter features. Then, the expectation maximization method and the morphological opening operation were used to obtain the yarn evenness. Finally, we calculated the yarn apparent evenness diameter and the coefficient of variation of the evenness of the yarn apparent diameter. Compared with the capacitive evenness testers, the Otsu detection method and the fuzzy C-means detection method, our method can accurately detect the yarn apparent evenness better than the selected state-of-the-art methods.
Draw textured yarn (DTY) packages is a significant raw material in manufacturing. Various defects will be generated on surface during production and transportation, of which hairiness is the most ...common and intractable defect. Many methods have been applied for fabric surface defect detection, but little research is aimed at DTY packages hairiness defects. In order to achieve the accuracy of DTY packages hairiness detection in industrial production, a method based on multi-directional anisotropic Gaussian directional derivatives was proposed to accomplish the DTY packages hairiness defect detection. Firstly, the original defect images were obtained by a device consisting of plane array light source, camera, and computer with image processing algorithm. Secondly, the gradient information of DTY packages images was constructed by anisotropic Gaussian directional derivative to characterize the defect. Then image response maps with all directions were fused to obtain the final response map. After that, a special difference of median (DOM) filter was proposed to remove useless information. Finally, the segmentation result was obtained by threshold method and morphological processing. Compared with various classical methods, the proposed method obtained the best performance in our evaluation experiments about DTY packages hairiness detection.
Example learning-based single image super-resolution (SR) technique has been widely recognized for its effectiveness in restoring a high-resolution (HR) image with finer details from a given ...low-resolution (LR) input. However, most popular approaches only choose one type of image features to learn the mapping relationship between LR and HR images, making it difficult to fit into the diversity of different natural images. In this paper, we propose a novel stacking learning-based SR framework by extracting both the gradient features and the texture features of images simultaneously to train two complementary models. Since the gradient features are helpful to represent the edge structures while the texture features are beneficial to restore the texture details, the newly proposed method cleverly combines the merits of two complementary features and makes the resultant HR images more faithful to their original counterparts. Moreover, we enhance the SR capacity by using a residual cascaded scheme to further reduce the gap between the super-resolved images and the corresponding original images. Experimental results carried out on seven benchmark datasets indicate that the proposed SR framework performs better than other seven state-of-the-art SR methods in both quantitative and qualitative quality assessments.
PDF
