Review: Application of Convolutional Neural Network in Defect Detection of 3C Products

Ming, Wuyi; Cao, Chen; Zhang, Guojun; Zhang, Hongmei; Zhang, Fei; Jiang, Zhiwen; Yuan, Jie

doi:10.1109/access.2021.3116131

Cited by 19 publications

(15 citation statements)

References 62 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…When dealing with MEA material structure optimization modeling in the future, reliability design should be integrated into the simulation process to improve the life of PEMFCs. (3) With the rapid development of computer technology, ML technology is widely used in the fields of industrial inspection and measurement [130][131][132], medical diagnosis [133,134], life sciences [135,136] and so on. For example, AlphaFold2 constructed a protein structure prediction model through ML; it was able to predict the properties of proteins from gene sequences, obtaining 98.5% of the human protein structure [135].…”

Section: Discussionmentioning

confidence: 99%

Review: Modeling and Simulation of Membrane Electrode Material Structure for Proton Exchange Membrane Fuel Cells

Chen

Liu

et al. 2022

Coatings

Self Cite

View full text Add to dashboard Cite

Hydrogen energy is recognized as the clean energy with the most development potential, and hydrogen fuel cell technology is considered the ultimate solution utilizing hydrogen energy. The proton exchange membrane fuel cell (PEMFC) has the merits of high energy efficiency, high energy density, low operating temperature, is clean, and affords environmental protection. Improving the structure of each functional layer could play a significant role in improving PEMFC performance. In addition, membrane electrode assemblies (MEAs) are the core components of a PEMFC, and their structure includes three main parts, namely, the gas diffusion layer (GDL), catalytic layer (CL), and proton exchange membrane (PEM). Therefore, this review focuses on progress in the modeling and simulation of the material structure in MEAs. First, the GDL simulation models are critically reviewed, including two-phase calculation models and microscopic simulation models. Second, CL microstructure models are comprehensively evaluated, involving power density enhancement, catalyst loading distribution, electrochemical reaction and its performance optimization. Third, the PEM simulation model, relating to molecular dynamics (MD) simulation techniques, 3D numerical techniques, and multiphysics simulation, are reviewed. Finally, the three aspects of similarity, individuality, and complementarity of these simulation models are discussed, and necessary outlooks, including the current limitations and challenges, are suggested, providing a reference for low-cost, high-performing PEMFC membrane electrodes for the future.

show abstract

Section: Discussionmentioning

confidence: 99%

Review: Modeling and Simulation of Membrane Electrode Material Structure for Proton Exchange Membrane Fuel Cells

Chen

Liu

et al. 2022

Coatings

Self Cite

View full text Add to dashboard Cite

show abstract

“…(4) The choice of classifier is one of the key points of Mura detection. Commonly used classification methods have certain advantages in one aspect, but their shortcomings are also observable in the other [82]. Therefore, combining the advantages of various classifiers and breaking through their limitations is one of the future research directions.…”

Section: Discussionmentioning

confidence: 99%

Survey of Mura Defect Detection in Liquid Crystal Displays Based on Machine Vision

Ming

Zhang

et al. 2021

Crystals

Self Cite

View full text Add to dashboard Cite

Liquid crystal display (LCD) is a display device based on liquid crystal electro-optic effect, and LCDs have gradually appeared and have become an indispensable part of people’s lives. In the development of LCD technology, the detection of Mura defects is a key concern in the manufacturing process. The Mura defect is a kind of display defect with low contrast and an irregular shape. This study first explains the mechanism of Mura defects in the LCD manufacturing process and classifies typical Mura defects. Then, three main purposes for the defect detection of LCDs are compared, and the advantages and disadvantages are conducted. Following that, this research examines reviews the linked literature on image preprocessing, feature extraction, dimension reduction, and classifiers of Mura defects. Finally, the future development trend and research direction of Mura defect detection based on machine vision can be drawn by this study.

show abstract

“…These algorithms will be improved as they are applied, and it is believed that future artificial intelligence will also produce more new algorithms that the optimization algorithm will apply to the EDM, such as selfish herds optimization, bald eagle search, etc. (4) With the fast advancement of technology, machine learning (ML) has found widespread use in a variety of industries, including industrial testing [121][122][123], medical diagnostics [124,125], life sciences [126,127], and renewable energy [128][129][130]. AlphaFold2, for instance, created a protein structure prediction model using ML, which can predict the properties of proteins based on gene sequences and achieve 98.5% of the structure of human proteins [126].…”

Section: Outlooksmentioning

confidence: 99%

Parameters Optimization of Electrical Discharge Machining Process Using Swarm Intelligence: A Review

Chen

et al. 2023

Metals

Self Cite

View full text Add to dashboard Cite

Electrical discharge machining (EDM) can use soft tool electrodes to process hard workpieces to achieve “soft against hard”, because it directly uses electrical energy and thermal energy to remove metal materials. Then, it can generate complex features on harder materials and meet the requirements of excellent surface quality. Since EDM involves many process parameters, including electrical parameters, non-electrical parameters, and materials properties, it is essential to optimize its process parameters to obtain good performance. In this direction, the application of the swarm intelligence (SI) technique has become popular. In this paper, the existing literature is comprehensively reviewed, and the application of the SI technique in the optimization of EDM process parameters is summarized. Sinker-EDM (SEDM), wire-EDM (WEDM), and micro-EDM (MEDM) with various hybrid techniques are among the EDM methods considered in this study because of their broad adoption in industrial sections. The fundamental nature of all review articles will assist engineers/workers in determining the process parameters and processing performance, the SI algorithm, and the optimal technique by which to obtain the desired process parameters. In addition, discussions from the perspectives of the similarity, individuality, and complementarity of various SI algorithms are proposed, and necessary outlooks are predicted, which provides references for the high performance of the EDM processes in the future.

show abstract

Review: Application of Convolutional Neural Network in Defect Detection of 3C Products

Cited by 19 publications

References 62 publications

Review: Modeling and Simulation of Membrane Electrode Material Structure for Proton Exchange Membrane Fuel Cells

Review: Modeling and Simulation of Membrane Electrode Material Structure for Proton Exchange Membrane Fuel Cells

Survey of Mura Defect Detection in Liquid Crystal Displays Based on Machine Vision

Parameters Optimization of Electrical Discharge Machining Process Using Swarm Intelligence: A Review

Contact Info

Product

Resources

About