Dual-camera framework for detecting subtle, low-contrast defects on multi-layered cover glass in display...

Area of Science:

• Materials Science
• Optical Engineering
• Computer Vision

Background:

• Smartphone display manufacturing requires stringent quality control for optical performance.
• Inspecting multi-layer transparent cover glass presents challenges like interference, low-contrast defects, and anisotropic noise.

Purpose of the Study:

• To develop a co-designed hardware-algorithm framework for accurate and efficient detection of cover glass defects.
• To address the limitations of post-lamination inspection in display manufacturing.

Main Methods:

• A dual-camera imaging system for multi-perspective synchronized acquisition to handle transparency interference.
• Attention-guided Fast-Wavelet Transform (FT-SWT) fusion for isolating low-contrast defect features.
• Anisotropic multi-scale Gaussian filtering to remove directional noise while preserving boundaries.

Main Results:

• The framework achieved 98.7% detection precision on industrial samples.
• Real-time detection capability was demonstrated.
• Successfully identified micron-scale particles, adhesive voids, and texture anomalies.

Conclusions:

• The proposed framework offers a robust solution for quality control of transparent multi-layer composites in display manufacturing.
• This methodology enhances defect detection accuracy and efficiency in smartphone production.