Automatic fabric defect detection using a wide-and-light network

Wu, Jun; Liu, Juan; Xiao, Zhitao; Zhang, Fang; Geng, Lei; Liu, Yanbei; Wang, Wen

doi:10.1007/s10489-020-02084-6

Cited by 45 publications

(15 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ntotal means the total number of windows. Detection accuracy is a benchmark criteria which is used most of classification problems in different scopes [25][26][27][28][29][30][31][32][33].…”

Section: Resultsmentioning

confidence: 99%

Visual Abnormality Detection in Surface based on Energy Variations in Multi Direction Co-occurrence Matrixes

Kiani¹

2022

Preprint

View full text Add to dashboard Cite

Surface defect detection is one of the most widely used research areas in the field of image processing and machine vision. Detection of surface defects is used in visual inspection systems and medical image analysis. In this manuscript, an innovative method for detecting surface defects based on energy changes in co-occurrence matrices is presented in several directions. The method presented in this manuscript includes two stages of learning and testing. In the learning phase, to extract texture features, the gray level co-occurrence matrix operator is applied on the healthy image of the desired level. Then the energy value of the output matrix is calculated. In the following, changes in the amount of energy are considered as statistical characteristics that are a good representative of the image of a healthy surface. Finally, with its help, a suitable threshold for the health of the images is obtained. Then, in the test phase, with the help of the calculated quorum, the defective windows that have suffered from non-normality are distinguished from the healthy surface sections. In the results section, the efficiency of the mentioned method has been measured on medical images and stone and ceramic images, and its detection accuracy has been compared with some previous effective methods. The advantages of the presented method include high accuracy, low calculations and compatibility with all types of levels due to the use of the learning stage. The proposed approach can be used in medical applications to diagnose abnormalities such as diseases. All extracted features are statistical, so its detection speed is higher than deep neural networks.

show abstract

Section: Resultsmentioning

confidence: 99%

Visual Abnormality Detection in Surface based on Energy Variations in Multi Direction Co-occurrence Matrixes

Kiani¹

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Hence, some research developed from the RCNN 21 has been adapted to fabric defect detection. 22 Li and Li 23 employed numerous tricks on the C-RCNN to improve the accuracy of textile defect detection, but there are no developments in the model structure or algorithmic strategy.…”

Section: Related Workmentioning

confidence: 99%

Fabric defect detection via a spatial cloze strategy

Zhang

Han

et al. 2022

Textile Research Journal

View full text Add to dashboard Cite

Deep-learning models have achieved state-of-the-art performances in a wide range of defect detection tasks. However, an inescapable criticism of one-stage fully supervised models is the lack of interpretability, which not only reduces the reliability of fabric defect detection systems but also limits the scope of their applications in production environments. To tackle the data imbalance and low interpretability of defect samples, we proposed a spatial cloze strategy for fabric defect detection, which reconstructs a local normal image and then feeds it into the detection model with the original image simultaneously. Specifically, we formulate the defect detection task as a novel image completion problem. Firstly, an end-to-end deep neural network is trained to finely restore the defect image by completing each image slice removed in sequence. Next, the progressive attention mechanism fuses the repaired normal image with the raw image, replacing the input layer of the cascade region-based convolutional neural network. Eventually, accurate instance-level defect segmentation can be obtained by comparing the repaired defect-free and the raw images. On the Tianchi dataset, the proposed method displays superior accuracy in 92% of defect classes, with a breakthrough in various categories that have hardly ever been detected. Extensive experiments on various complex fabric defect samples demonstrate that our strategy outperforms existing advanced methods.

show abstract

“…Published by Francis Academic Press, UK -82-of the small target. Jun Wu et al [5] applied the Faster R-CNN model to the defect detection of the fabric. They mainly designed a dilated convolution module, and got multi-scale features by connecting features map of different scales to improve the accuracy of fabric defect detection.…”

Section: Related Workmentioning

confidence: 99%

Fabric Defect Detection Based on the Improved Cascade R-CNN

2021

AJCIS

View full text Add to dashboard Cite

There is a large amount of fabric produced in the process of industrial production, thus fabric defects automatic detection can bring great benefits to enterprises. With the development of computer technology, deep learning has more advantages on fabric defect detection compared with traditional image processing. By comparing the advantages and disadvantages of different target detection models, we chose to use the Cascade R-CNN model for fabric defect detection finally. However, due to the large size of fabric image, small number of partial defects and complex image background, there are many difficulties in fabric defect detection. To solve the difficulties, we propose some methods to improve the accuracy of defect detection. First, in order to enhance the detection accuracy of small targets, we cut large fabric images into blocks for training and detection, and combine the detection results. Then, to solve the problem of the small number of partial defects in the fabric Dataset, a new method of multi-morphological data augmentation was proposed to increase the number of the Dataset. Finally, we improved the Feature Pyramid Networks module to enhance positional accuracy of defect detection. Experimental results show that the model can complete the task of defect detection efficiently and improve the accuracy of target detection effectively.

show abstract

Automatic fabric defect detection using a wide-and-light network

Cited by 45 publications

References 30 publications

Visual Abnormality Detection in Surface based on Energy Variations in Multi Direction Co-occurrence Matrixes

Visual Abnormality Detection in Surface based on Energy Variations in Multi Direction Co-occurrence Matrixes

Fabric defect detection via a spatial cloze strategy

Fabric Defect Detection Based on the Improved Cascade R-CNN

Contact Info

Product

Resources

About