Directional textures auto-inspection using principal component analysis

Chen, Ssu-Han; Perng, Der‐Baau

doi:10.1007/s00170-010-3141-1

Cited by 29 publications

(16 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Low-rank-approximation-based methods reconstruct the regular textured background using a low-rank approximation of the given image under the assumption that the textured background is well-approximated using a low-rank matrix. Some methods first reconstruct the textured background and subtract the reconstructed background from the input image [7,11]. Other methods attempted to directly estimate the foreground defect image using a low-rank approximation [8][9][10].…”

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

“…Other methods attempted to directly estimate the foreground defect image using a low-rank approximation [8][9][10]. After removing the textured background, some methods relied on the SPC to detect the defective region [8][9][10][11]19], and another method used the binarization of the background-removed image to find the defective region [7]. To reconstruct either the foreground defect image directly or to reconstruct the textured background image, low-rank approximation methods such as SVD [8,9], PCA [11] and ICA [7] have been applied.…”

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

“…After removing the textured background, some methods relied on the SPC to detect the defective region [8][9][10][11]19], and another method used the binarization of the background-removed image to find the defective region [7]. To reconstruct either the foreground defect image directly or to reconstruct the textured background image, low-rank approximation methods such as SVD [8,9], PCA [11] and ICA [7] have been applied. In addition, in one study, DCT was applied and several coefficients related to the foreground images were obtained to reconstruct the foreground image [10].…”

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

“…Some previous investigations attempted to reconstruct the textured background and subtract the background from the acquired image [8][9][10]. Other studies attempted to estimate the foreground defects directly from the acquired image [7,11]. Both of these types of methods relied on low-rank approximation techniques such as singular value decomposition (SVD), principal component analysis (PCA), and independent component analysis (ICA) [7][8][9]11], which will be explained detail in Section 2.1.…”

Section: Introductionmentioning

confidence: 99%

“…The existing methods usually apply the defect detection method after extracting the defective parts of image. In some studies, a statistical process control (SPC) method that detects defective pixels based on the variation from the mean of the image was applied [8][9][10][11], while in an another study, simple binarization was applied for detecting the defective regions [7].…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Regularized Auto-Encoder-Based Separation of Defects from Backgrounds for Inspecting Display Devices

Kim

2019

Electronics

View full text Add to dashboard Cite

We investigated a novel method for separating defects from the background for inspecting display devices. Separation of defects has important applications such as determining whether the detected defects are truly defective and the quantification of the degree of defectiveness. Although many studies on estimating patterned background have been conducted, the existing studies are mainly based on the approach of approximation by low-rank matrices. Because the conventional methods face problems such as imperfect reconstruction and difficulty of selecting the bases for low-rank approximation, we have studied a deep-learning-based foreground reconstruction method that is based on the auto-encoder structure with a regression layer for the output. In the experimental studies carried out using mobile display panels, the proposed method showed significantly improved performance compared to the existing singular value decomposition method. We believe that the proposed method could be useful not only for inspecting display devices but also for many applications that involve the detection of defects in the presence of a textured background.

show abstract

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

Section: Low-rank-approximation-based Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Regularized Auto-Encoder-Based Separation of Defects from Backgrounds for Inspecting Display Devices

Kim

2019

Electronics

View full text Add to dashboard Cite

show abstract

Liquid crystal display defects in multiple backgrounds with visual real‐time detection

Cui

Wang

et al. 2021

J Soc Info Display

View full text Add to dashboard Cite

There are kinds of defects that may appear in the process of Liquid Crystal Display (LCD) manufacturing, which cannot be effectively detected, owing to the uneven illumination, low contrast, and miscellaneous patterns of defects. To improve the efficiency of defect detection and ensure the quality of LCD, three visual real‐time detection methods are adopted for detecting six different defects in multiple backgrounds, where image preprocessing methods are used to highlight the defects and facilitate the segmentation and detection. Specifically, the interclass variance (OTSU) method is used to segment and mark Liquid Crystal Display (LCD) Mura and scratch defects in six kinds of solid color backgrounds; the method and the connectivity‐4 judgment criteria are adopted to label edge defects in grid display background; the gray mean and standard deviation of the segmented subregions are calculated to recognize the color gradation defect in the 32‐level gradation display background. Experimental results show that LCD Mura defects and scratches can be segmented more completely by the proposed method compared with the benchmark methods, and the edge defects can be identified accurately by the OTSU‐based method and particle‐based morphological processing with grids as the detection background, and the color gradation can also be recognized with the 32‐level gray gradation as the background.

show abstract

Automated inspection of engineering ceramic grinding surface damage based on image recognition

Chen

Lin

Han

et al. 2012

Int J Adv Manuf Technol

View full text Add to dashboard Cite

Directional textures auto-inspection using principal component analysis

Cited by 29 publications

References 12 publications

Regularized Auto-Encoder-Based Separation of Defects from Backgrounds for Inspecting Display Devices

Regularized Auto-Encoder-Based Separation of Defects from Backgrounds for Inspecting Display Devices

Liquid crystal display defects in multiple backgrounds with visual real‐time detection

Automated inspection of engineering ceramic grinding surface damage based on image recognition

Contact Info

Product

Resources

About