Waveguide quality inspection in quantum cascade lasers: A capsule neural network approach
作者：
Highlights：
• A deep learning model to classify anomalies in optoelectronic wafers.
• Defects and dirt accumulated in waveguide were the two anomalies investigated.
• The deep learning approach combines CNN and WaferCaps by parallel decision fusion.
• Image data for analysis were collected from laser manufacturer in Europe.
• The results showed the importance of using the decision fusion approach.
摘要
•A deep learning model to classify anomalies in optoelectronic wafers.•Defects and dirt accumulated in waveguide were the two anomalies investigated.•The deep learning approach combines CNN and WaferCaps by parallel decision fusion.•Image data for analysis were collected from laser manufacturer in Europe.•The results showed the importance of using the decision fusion approach.
论文关键词：Automatic optical inspection,Capsule networks,Convolutional neural networks,Deep learning,Defect inspection,Optoelectronic industry,Quantum cascade lasers
论文评审过程：Received 21 November 2021, Revised 16 June 2022, Accepted 3 August 2022, Available online 11 August 2022, Version of Record 17 August 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118421
