Defect Classification of Printed Circuit Boards based on Transfer Learning

Ghosh, Barnajit; Bhuyan, M. K.; Sasmal, Pradipta; Iwahori, Yuji; Gadde, Prathik

doi:10.1109/aspcon.2018.8748670

Cited by 27 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Ultimately, they achieved a detection accuracy of 98.4% on the DeepPCB dataset using only 50 labeled samples. Ghosh et al, performed transfer learning using a pre-trained Inception-V3 model [5]. They extracted the mid-level neural network of Inception-V3 as the front half of the defect detection network and an adaptation network as the back half of the detection network.…”

Section: Solutionmentioning

confidence: 99%

Research on the application of deep learning algorithms to PCB defect detection

Kong

2024

ACE

View full text Add to dashboard Cite

In today's electronics industry, Printed Circuit Boards play a crucial role in providing the layout for circuit components and conductive traces in nearly all electronic devices. The quality of components soldered onto Printed Circuit Boards directly impacts product performance. To ensure the performance of electronic devices, Printed Circuit Boards defect detection based on deep learning algorithms has become a pivotal technology in the defect inspection process within the electronics industry. However, the application of deep learning algorithms in this context faces several challenges. These challenges include difficulties in acquiring Printed Circuit Boards defect datasets, limited generalization capability in Printed Circuit Boards defect detection, and slow and low-quality Printed Circuit Boards image stitching processes. To enhance researchers' understanding of deep learning-based Printed Circuit Boards defect detection, this paper analyzes the challenges associated with deep learning in the Printed Circuit Boards defect detection process and proposes several viable solutions. In conclusion, this paper provides insights into the future of deep learning-based Printed Circuit Boards defect detection.

show abstract

Section: Solutionmentioning

confidence: 99%

Research on the application of deep learning algorithms to PCB defect detection

Kong

2024

ACE

View full text Add to dashboard Cite

show abstract

“…In stacked autoencoder, the cross entropy loss function is applied to deal with the optimization problem, which is mathematically determined in Eq. (12).…”

Section: Dimensionality Reductionmentioning

confidence: 99%

“…In addition to this, a high level multiple dimensional classification techniques such as random forest, support vector machine, etc. are employed for identifying the defect samples [12]. Whereas, the success of the classification techniques completely depends on the human experts for extracting and selecting the representative feature values based on the structure variations and local gray level variations of a defect in the test PCB defective image [13].…”

Section: Introductionmentioning

confidence: 99%

Effective PCB Defect Detection Using Stacked Autoencoder with Bi-LSTM Network

2022

IJIES

View full text Add to dashboard Cite

In the Printed Circuit Boards (PCBs) manufacturing, the defect detection is an important task that helps in improving the quality of PCBs production. The conventional defect detection schemes include few drawbacks like high computational cost, noise susceptibility, and strongly depends on a carefully designed template. To highlight the above stated drawbacks, a new hybrid deep learning model is proposed in this manuscript. Initially, the input PCB defective images are collected from the PCB defect dataset and further, the feature extraction is performed by utilizing Binary Robust Invariant Scalable Keypoints (BRISK) and Speeded up Robust Features (SURF) for extracting the feature vectors from the PCB defective image. The extracted feature vectors are multi-dimensional that increase the computational complexity, so the stacked autoencoder technique is applied for reducing dimension of the extracted feature vectors. The stacked autoencoder technique effectively selects minimal sub-set of non-redundant and relevant feature vectors and it is used for representing the datasets from original feature space to a reduced and more informative feature space. Finally, the selected feature vectors are fed to the Bi-Long Short Memory Network (Bi-LSTM) for classifying the defect types like mouse bite, spurious copper, short, spur, missing hole, and open circuit. The extensive experimental outcome confirmed that the hybrid deep learning model obtained higher performance in the PCB defect detection with the classification accuracy of 99.99%, which is superior related to the comparative models.

show abstract

Inspection of Lead Frame Defects Using Deep CNN and Cycle-Consistent GAN-Based Defect Augmentation

Juang

Chen

2022

Intelligent Systems Reference Library

View full text Add to dashboard Cite

Defect Classification of Printed Circuit Boards based on Transfer Learning

Cited by 27 publications

References 7 publications

Research on the application of deep learning algorithms to PCB defect detection

Research on the application of deep learning algorithms to PCB defect detection

Effective PCB Defect Detection Using Stacked Autoencoder with Bi-LSTM Network

Inspection of Lead Frame Defects Using Deep CNN and Cycle-Consistent GAN-Based Defect Augmentation

Contact Info

Product

Resources

About