Surface defect detection method for air rudder based on positive samples

Yang, Zeqing; Zhang, Mingxuan; Chen, Yingshu; Hu, Ning; Gao, Lingxiao; Liu, Libing; Ping, Enxu; Song, Jung Il

doi:10.1007/s10845-022-02034-8

Cited by 7 publications

(3 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Object boundaries, edges and other details are high frequency components of the image; usually defects, like ''scratch'', are recognizable among these categories. Yang et al [80] adopted a Frequency-shifted convolutional layer to tackle high frequency information loss at the expense of semantic information prevalence in deeper layers. Other mathematical operations are being explored, such as Atrous (or Deformable) Convolution which helps combine sparse encoded information, by connecting feature related to non-adjacent image regions [81].…”

Section: ) Conceptualizationmentioning

confidence: 99%

“…2) introducing stochastic variations or modifying lighting conditions: it consists of adding Gaussian noise, random brightness changing, enhancing contrast [74], [91], [95], [96], [97], [98], [99], using circular or elliptical templates [90]. 3) generative models: they include Conditional-Convolutional Variational AE [65], [100] and Deep Convolutional GAN [80].…”

Section: A: Overcoming Imbalanced Datamentioning

confidence: 99%

“…The method is built on a convolutional-GAN that helps establish a decision boundary robust to anomaly data infiltration since it has purpose of optimizing the normal image space by recovering from latent space. Yang et al [80] introduced a reconstruction error based on residuals to quantify a pooled abnormal image pattern score, which, once compared with a threshold, generated a binary class label. Niu et al [45] adopted a memory block to preserve historic information and to learn the group characteristics of the defect-free samples.…”

Section: B: Reconstruction-based Approachesmentioning

confidence: 99%

See 2 more Smart Citations

Deep Learning for Automatic Vision-Based Recognition of Industrial Surface Defects: A Survey

et al. 2023

View full text Add to dashboard Cite

Automatic vision-based inspection systems have played a key role in product quality assessment for decades through the segmentation, detection, and classification of defects. Historically, machine learning frameworks, based on hand-crafted feature extraction, selection, and validation, counted on a combined approach of parameterized image processing algorithms and explicated human knowledge. The outstanding performance of deep learning (DL) for vision systems, in automatically discovering a feature representation suitable for the corresponding task, has exponentially increased the number of scientific articles and commercial products aiming at industrial quality assessment. In such a context, this article reviews more than 220 relevant articles from the related literature published until February 2023, covering the recent consolidation and advances in the field of fully-automatic DL-based surface defects inspection systems, deployed in various industrial applications. The analyzed papers have been classified according to a bi-dimensional taxonomy, that considers both the specific defect recognition task and the employed learning paradigm. The dependency on large and high-quality labeled datasets and the different neural architectures employed to achieve an overall perception of both well-visible and subtle defects, through the supervision of fine or/and coarse data annotations have been assessed. The results of our analysis highlight a growing research interest in defect representation power enrichment, especially by transferring pre-trained layers to an optimized network and by explaining the network decisions to suggest trustworthy retention or rejection of the products being evaluated.

show abstract

Section: ) Conceptualizationmentioning

confidence: 99%