Explore the vast range of titles available.

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.

To improve the detection rate of defect and the fabric product quality, a higher real-time performance fabric defect detection method based on the improved YOLOv3 model is proposed. There are two key steps: first, on the basis of YOLOv3, the dimension clustering of target frames is carried out by combining the fabric defect size and k-means algorithm to determine the number and size of prior frames. Second, the low-level features are combined with the high-level information, and the YOLO detection layer is added on to the feature maps of different sizes, so that it can be better applied to the defect detection of the gray cloth and the lattice fabric. The error detection rate of the improved network model is less than 5% for both gray cloth and checked cloth. Experimental results show that the proposed method can detect and mark fabric defects more effectively than YOLOv3, and effectively reduce the error detection rate.

To solve the problem of defect detection in printed fabrics caused by abundant colors and varied patterns, a defect detection method based on RGB accumulative average method (RGBAAM) and image pyramid matching is proposed. First, the minimum period of the printed fabric is calculated by the RGBAAM. Second, a Gaussian pyramid is constructed for the template image and the detected image by using the minimum period as a template. Third, the similarity measurement method is used to match the template image and the detected image. Finally, the position of the printed fabric defect is marked in the image to be detected by using the Laplacian pyramid restoration. The experimental results show that the method can accurately segment the printed fabric periodic unit and locate the defect position. The calculation cost is low for the method proposed in this article.

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.

Objective To explore the value of three-dimensional visual reconstruction technology and virtual diagnosis and treatment technology in robot-assisted laparoscopic partial nephrectomy (RAPN) for complex renal tumors (R.E.N.A.L. score ≥ 7). Methods A retrospective analysis was conducted on the clinical data of 60 patients scheduled to undergo RAPN for complex renal tumors at Hefei Second People’s Hospital between January 2021 and January 2025 and confirmed by postoperative pathological examination. The patients were divided into two groups: the 3D group ( n  = 30), which underwent three-dimensional visual reconstruction and virtual diagnosis and treatment technology for preoperative planning, and the 2D imaging group ( n  = 30), which underwent preoperative planning using enhanced CT and enhanced MRI. The following parameters were compared between the two groups: preoperative data (gender, age, body mass index, tumor diameter, R.E.N.A.L. score, laterality, and tumor location); intraoperative and postoperative indicators (total operation time, warm ischemia time, intraoperative blood loss, hemoglobin change, postoperative hospital stay, duration of drain retention, total cost, and blood transfusion rate); renal function change indicators (changes in serum creatinine and glomerular filtration rate preoperatively and at 1, 3, and 6 months postoperatively); and surgical complications (bleeding, subcutaneous emphysema, urinary leakage, intestinal injury). Results There were no significant differences in the baseline characteristics (including age, gender, BMI, tumor laterality, tumor diameter, and R.E.N.A.L. score) between the two groups( P >0.05). Regarding surgical parameters, the 3D reconstruction group showed superior outcomes compared to the 2D imaging group, with statistically significant differences observed in the following metrics༈ P <0.05༉: operative time (135.70 ± 18.41 min vs. 142.37 ± 14.25 min, P  = 0.025), warm ischemia time (24.67 ± 5.48 min vs. 28.07 ± 5.92 min, P  = 0.013), intraoperative blood loss (85.03 ± 25.23 mL vs. 99.67 ± 30.03 mL, P  = 0.025), and hemoglobin change (6.97 ± 2.48 g/L vs. 8.67 ± 3.26 g/L, P  = 0.022). The 3D reconstruction group showed favorable, though not statistically significant༈ P >0.05༉, trends in the following parameters compared to the 2D imaging group: tumor resection time (11.27 ± 2.77 min vs. 12.80 ± 3.73 min, P  = 0.059), postoperative hospital stay (7.70 ± 1.91 days vs. 8.57 ± 2.03 days, P  = 0.063), time to drain removal (5.83 ± 1.15 days vs. 6.23 ± 1.17 days, P  = 0.159), change in serum creatinine levels (11.20 ± 4.28 µmol/L vs. 12.53 ± 4.22 µmol/L), and change in glomerular filtration rate (14.33 ± 3.13 mL/min vs. 16.07 ± 3.77 mL/min). Regarding postoperative complications, no statistically significant difference was found between the two groups༈ P >0.05༉. In the 3D group, complications included 2 cases of bleeding and 1 case of subcutaneous emphysema, with no instances of urinary leakage or intestinal injury. The 2D group had 1 case of urinary leakage, 2 cases of bleeding, 1 case of subcutaneous emphysema, and 2 cases of intestinal injury. In the 3D group, one patient with postoperative urinary fistula was classified as Clavien-Dindo grade IIIa, managed by ureteral stent placement under local anesthesia. Two patients with postoperative hemorrhage in the 3D group and two in the 2D group were classified as Clavien-Dindo grade IIIa, treated with arterial embolization in the interventional department. One patient in the 3D group and one in the 2D group were classified as Clavien-Dindo grade I, managed by clinical observation. One patient with intestinal injury in the 2D group was classified as Clavien-Dindo grade II, treated with antibiotic therapy for infection control. In the long-term renal function change indicators of the two groups of patients, data collection was limited to the 1st, 3rd, and 6th months post-surgery due to the follow-up period constraints. The changes in serum creatinine levels at 1 month postoperatively in the 3D reconstruction group (12.67 ± 4.40 µmol/L vs. 13.07 ± 3.82 µmol/L, P  = 0.543), the changes in glomerular filtration rate at 1 month postoperatively (13.20 ± 3.07 ml/min vs. 14.70 ± 3.48 ml/min, P  = 0.079), the changes in serum creatinine levels at 3 months postoperatively (7.60 ± 3.58 µmol/L vs. 8.50 ± 4.81 µmol/L, P  = 0.467), the changes in glomerular filtration rate at 3 months postoperatively (7.00 ± 2.32 ml/min vs. 8.60 ± 3.28 ml/min, P  = 0.053), the changes in serum creatinine levels at 6 months postoperatively (8.13 ± 1.70 µmol/L vs. 9.13 ± 2.71 µmol/L, P  = 0.155), and the changes in glomerular filtration rate at 6 months postoperatively (7.17 ± 2.68 ml/min vs. 7.83 ± 2.57 ml/min, P  = 0.482) showed no statistically significant differences༈ P >0.05༉. Conclusion The application of three-dimensional visual reconstruction and digital virtual diagnosis and treatment technology in RAPN for complex renal tumors can reduce operative time and warm ischemia time, decrease intraoperative blood loss, and shows certain value in enhancing surgical safety. Furthermore, the use of 3D visualization technology improves surgical efficiency without increasing the rate of postoperative complications, representing a safe and effective methodology.

In this paper, we present a robust and reliable framework based on L0 gradient minimization (LGM) and the fuzzy c-means (FCM) method to detect various fabric defects with diverse textures. In our framework, the L0 gradient minimization is applied to process the fabric images to eliminate the influence of background texture and preserve sharpened significant edges on fabric defects. Then, the processed fabric images are clustered by using the fuzzy c-means. Through continuous iterative calculation, the clustering centers of fabric defects and non-defects are updated to realize the defect regions segmentation. We evaluate the proposed method on various samples, which include plain fabric, twill fabric, star-patterned fabric, dot-patterned fabric, box-patterned fabric, striped fabric and statistical-texture fabric with different defect types and shapes. Experimental results demonstrate that the proposed method has a good detection performance compared with other state-of-the-art methods in terms of both subjective and objective tests. In addition, the proposed method is applicable to industrial machine vision detection with limited computational resources.

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.

In order to solve the problem of unstable quality of the traditional bobbin yarn surface defect detection, a method for detecting the surface defect of bobbin yarn using visual saliency analysis is proposed. Firstly, the Difference of Gaussian algorithm is used to suppress the texture of the image, and the contrast between the defect area and the background is enhanced. Then, the method of wavelet threshold is applied to filter the noise in the images after the Difference of Gaussian, and the detailed features of the defect area are preserved while the noise interference is eliminated. Finally, the frequency tuned visual saliency algorithm is used to extract and segment the defective regions accurately. After the process about the Difference of Gaussian and wavelet threshold denoising, the original image of the bobbin yarn can make the image highlight the feature information of the defect area on the basis of ensuring elimination of the background texture and noise, and facilitate the detection of the defect area by combining visual saliency analysis. In the detection of the accuracy and integrity of the defect area, the detection method of this paper is better than the control method, and the defect segmentation result is more accurate. The proposed method for detecting the surface defect of the bobbin can not only completely eliminate the influence of the background texture of the bobbin surface, but also accurately and completely detect the surface defect of the bobbin, and the defect details remain intact. In this paper, the effectiveness of the algorithm is further verified by objective tests. The precision and recall rate of the test are 95.15 % and 98.00 %, respectively, which can meet the needs of actual detection.

Fabric defect detection has always been a key issue, and it positively correlated its efficiency with productivity. From manual visual methods to machine vision and deep learning-based techniques, a variety of methods have been studied to improve production efficiency and product quality. Although deep learning-based methods have proven to be powerful tools for segmentation, there are still many pressing issues that need to be addressed in practical applications. First, the scarcity of defective samples compared to normal samples can cause data imbalance and thus affect accuracy. Second, high real-time performance is also required in the actual detection process. To overcome these problems, we propose a high real-time convolutional neural network, named Mobile-Deeplab, to implement end-to-end defect segmentation. In addition, we proposed a loss function to consider the fabric image sample imbalance problem. We evaluated the performance of the model with two public structured datasets and three self-constructed structured datasets. The experimental results show that the segmentation method has better segmentation accuracy than other segmentation models, which verifies the segmentation effect of the method. In addition, 87.11 frames per second on a 256×256 size image meet industrial real-time requirements.

The productivity of textile industry is positively correlated with the efficiency of fabric defect detection. Traditional manual detection methods have gradually been replaced by deep learning algorithms based on cloud computing due to the low accuracy and high cost of manual methods. Nonetheless, these cloud computing-based methods are still suboptimal due to the data transmission latency between the end devices and the cloud. To facilitate defect detection with more efficiency, a low-latency, low power consumption, easy upgrade, and automatical visual inspection system with the help of edge computing are proposed in this work. Firstly, the method uses EfficientDet-D0 as the detection algorithm, integrating the advantages of lightweight and scalable and can suit the resource-constrained edge device. Secondly, we performed data augmentations on five fabric datasets and verified the adaptability of the model in different types of fabrics. Finally, we transplanted the trained model to the edge device NVIDIA Jetson TX2 and optimized the model with TensorRT to make it detection faster. The performance of the proposed method is evaluated in five fabric datasets. The detection speed is up to 22.7 frame per second (FPS) on the edge device Jetson TX2. Compared with the cloud-based method, the response time is reduced by 2.5 times, with the capability of real-time industrial defect detection.