Effective Defect Detection Method Based on Bilinear Texture Features for LGPs

Automatic defect detection of light guide plates (LGPs) is an important task in the manufacture of liquid crystal displays. During thermo-printing, defects of tag lines on LGPs may occur easily, and these defects are of two categories: bubbles and missing tag lines. These defects lack salient visual...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Libin Hong, Xianglei Wu, Dibin Zhou, Fuchang Liu
Formato: article
Lenguaje:EN
Publicado: IEEE 2021
Materias:
Acceso en línea:https://doaj.org/article/2836c72c54954418a6ed8e0421d98d93
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
Descripción
Sumario:Automatic defect detection of light guide plates (LGPs) is an important task in the manufacture of liquid crystal displays. During thermo-printing, defects of tag lines on LGPs may occur easily, and these defects are of two categories: bubbles and missing tag lines. These defects lack salient visual attributes, such as edge-based and region-based features, and as such, traditional methods fail to detect them. To address this, we propose a Dense-bilinear convolutional neural network (BCNN), an end-to-end defect detection network, utilizing Dense-blocks (Huang <italic>et al.</italic>, 2017), Bilinear feature layers (Lin <italic>et al.</italic>, 2015), and squeeze-and-excitation blocks (Hu <italic>et al.</italic>, 2018). Our network exploits fine-grained texture features, which leads to parameter reduction and accuracy enhancement. We validate our network on our LGP dataset containing 5,860 images from three cases: bubbles, tag line existence, and tag line missing. Our network outperforms AlexNet (Krizhevsky <italic>et al.</italic>, 2012), VGG (Simonyan and Zisserman, 2014) and ResNet (He <italic>et al.</italic>, 2016), on both the public and our LGP datasets with less GPU memory consumption.