Detection of Monocrystalline Silicon Wafer Defects Using Deep Transfer Learning

Detection of Monocrystalline Silicon Wafer Defects Using Deep Transfer Learning

Defect detection is an important step in industrial production of monocrystalline silicon. Through the study of deep learning, this work proposes a framework for classifying monocrystalline silicon wafer defects using deep transfer learning (DTL). An existing pre-trained deep learning model was used...

Full description

Saved in:
Bibliographic Details
Published in:Journal of Telecommunications and Information Technology no. 1; pp. 34 - 42
Main Authors: Ganum, Adriana, Iskandar, D N F Awang, Chin, Lim Phei, Fauzi, Ahmad Hadinata
Format: Journal Article
Language:English
Published: Warsaw Instytut Lacznosci - Panstwowy Instytut Badawczy (National Institute of Telecommunications) 2022
National Institute of Telecommunications
Subjects:
Accuracy
automated optical inspection
Automation
Classification
Datasets
Deep learning
Defects
Feature extraction
imperfection identification
Machine learning
Manufacturers
Manufacturing
Microcracks
Model accuracy
neural network
Pattern recognition
Silicon
Silicon wafers
Training
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Defect detection is an important step in industrial production of monocrystalline silicon. Through the study of deep learning, this work proposes a framework for classifying monocrystalline silicon wafer defects using deep transfer learning (DTL). An existing pre-trained deep learning model was used as the starting point for building a new model. We studied the use of DTL and the potential adaptation of MobileNetV2 that was pre-trained using ImageNet for extracting monocrystalline silicon wafer defect features. This has led to speeding up the training process and to improving performance of the DTL-MobileNetV2 model in detecting and classifying six types of monocrystalline silicon wafer defects (crack, double contrast, hole, microcrack, saw-mark and stain). The process of training the DTL-MobileNetV2 model was optimized by relying on the dense block layer and global average pooling (GAP) method which had accelerated the convergence rate and improved generalization of the classification network. The monocrystalline silicon wafer defect classification technique relying on the DTL-MobileNetV2 model achieved the accuracy rate of 98.99% when evaluated against the testing set. This shows that DTL is an effective way of detecting different types of defects in monocrystalline silicon wafers, thus being suitable for minimizing misclassification and maximizing the overall production capacities.
ISSN:1509-4553
1899-8852
DOI:10.26636/jtit.2022.156321