Fast Surface Defect Detection Using Improved Gabor Filters

Ma, Jiaxu; Wang, Yuxi; Chen, Shi; Lu, Cewu

doi:10.1109/icip.2018.8451351

Cited by 17 publications

(9 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For surface defect detection approach, many researchers had worked on the problem and proposed different solutions since the last decades. The traditional two-dimensional vision algorithm uses a series of algorithms such as threshold [9] , lter [10] , morphology [11] to analyze the image, and uses the geometric shape and gray difference of the defects to recognize and classify the defects. The following scholars have contributed to these eld.…”

Section: Surface Defect Detection Approachmentioning

confidence: 99%

A novel approach for surface defectdetection of lithium battery based on improved K-Nearest Neighbor and Euclidean clustering segmentation

Liu¹,

Wu²,

Guo³

et al. 2023

Preprint

View full text Add to dashboard Cite

Surface defects of lithium batteries seriously affect the product quality and may lead to safety risks. In order to accurately identify the surface defects of lithium battery, a novel defect detection approach is proposed based on improved K-Nearest Neighbor (KNN) and Euclidean clustering segmentation. Firstly, an improved Voxel density strategy for KNN is proposed to speed up the effect for point filtering. Then, the improved clustering segmentation strategy is applied to distinguish point clouds with defect features. The outline fitting algorithm based on least square method is applied to determine geometric features of each surface defects. Furthermore, experimental results show that the proposed surface defect detection method reaches 99.2% accuracy and 35.3ms average time consumption for data processing. Finally, an industrial application example of lithium battery production is demonstrated, which meets the requirements of industrial application.

show abstract

Section: Surface Defect Detection Approachmentioning

confidence: 99%

A novel approach for surface defectdetection of lithium battery based on improved K-Nearest Neighbor and Euclidean clustering segmentation

Liu¹,

Wu²,

Guo³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…In addition, the hyperparameter α used for the Beta distribution in equation (9) and equation 12is set to 0.75. The hyperparameter α acts as a regulator to adjust the proportions of the two items LMSE and LKL in LU in equation (11), and the proportions of the two items LS and LU in loss in equation (13).…”

Section: Experiments and Comparisonsmentioning

confidence: 99%

“…Statistical methods utilize first-order or second-order statistics to extract defect features. Popular statistical methods include histogram properties [4], co-occurrence matrix [5] and local binary pattern [6].The spectral methods transform signals from the spatial domain to the frequency domain for defect identification through Fourier transform [7], wavelet transform [8] and Gabor filters [9]. Model-based methods capture features and identify defects by constructing an image model.…”

Section: Introductionmentioning

confidence: 99%

A Generic Semi-Supervised Deep Learning-Based Approach for Automated Surface Inspection

et al. 2020

View full text Add to dashboard Cite

Automated surface inspection (ASI) is critical to quality control in industrial manufacturing processes. Recent advances in deep learning have produced new ASI methods that automatically learn highlevel features from training samples while being robust to changes and capable of detecting different types of surfaces and defects. However, they usually rely heavily on manpower to collect and label training samples. In this paper, a generic semi-supervised deep learning-based approach for ASI that requires a small quantity of labeled training data is proposed. While the approach follows the MixMatch rules to conduct sophisticated data augmentation, we introduce a new loss function calculation method and propose a new convolutional neural network based on a residual structure to achieve accurate defect detection. An experiment on two public datasets (DAGM and NEU) and one industrial dataset (CCL) is carried out. For public datasets, the experimental results are compared against several best benchmarks in the literature. For the industrial dataset, the results are compared against deep learning methods based on benchmark neural networks. The proposed method achieves the best performance in all comparisons. In addition, a comparative experiment of model performance given a different number of labeled samples is conducted, demonstrating that the proposed method can achieve good performance with few labeled training samples.

show abstract

“…With the rapid development of modern optical technology, precision optical components have been widely used in some high-tech industries, and the accuracy requirements of optical components have increasingly improved. Since the various defects that are distributed randomly on the optical surface will obviously influence the application of the components, then the detection of surface defects has great necessity [1][2][3].…”

Section: Introductionmentioning

confidence: 99%

Influence of light polarization state on the imaging quality of dark-field imaging system

Chen¹,

Wang²,

Zhang³

2022

J. Opt.

View full text Add to dashboard Cite

Annular linear polarized light is used as the illumination source of the reflective dark-field detecting system in this paper. According to the theories of the Bidirectional Reflectance Distribution Function (BRDF) and multi-beam interference, the influence of the light polarization state on the intensity distribution of the scattering light is analyzed in detail. For surface defects, a simulation model of dark-field imaging is established based on the Finite-Difference Time-Domain method (FDTD). Both the near-field and the far-field scattering intensity distribution caused by surface defects are calculated under different illumination conditions. The incidence angle and polarization state of illumination light are optimized. Simulation and experimental results show that the image quality will be minimally affected by the interference effect while P-polarized light illuminates with the incident angle of 45°. The higher measurement accuracy of the dark-field imaging detection system can be obtained when the optimized illumination scheme is used.

show abstract

Fast Surface Defect Detection Using Improved Gabor Filters

Cited by 17 publications

References 17 publications

A novel approach for surface defectdetection of lithium battery based on improved K-Nearest Neighbor and Euclidean clustering segmentation

A novel approach for surface defectdetection of lithium battery based on improved K-Nearest Neighbor and Euclidean clustering segmentation

A Generic Semi-Supervised Deep Learning-Based Approach for Automated Surface Inspection

Influence of light polarization state on the imaging quality of dark-field imaging system

Contact Info

Product

Resources

About