Automated Optical Inspection for iPhone User Side Glass Logo
Segmentation-based inspection · Logo region defects · Reflective surface handling
Overview
Specialized inspection system for user-side glass logo area with precision segmentation to detect subtle defects. The system handles challenging reflective and low-contrast logo regions with robust defect localization and classification.
Problem
- Logo region has reflective/low-contrast characteristics.
- Subtle defects difficult to detect with traditional methods.
- Need for precise localization in small logo area.
- Variations in lighting and viewing angles affect detection.
Solution
- Segmentation-based inspection for reflective/low-contrast regions.
- Improved defect localization and classification robustness.
- Custom training with augmentation for various lighting conditions.
- Production-ready optimization and validation.
My Role
- Designed segmentation-based inspection for reflective/low-contrast logo regions.
- Improved defect localization and classification robustness.
- Production-ready optimization and validation.
- Tuned model for stable performance across manufacturing variations.
Implementation
Language: Python
Libraries:
PyTorch
OpenCV
Approach: Deep learning segmentation for logo-region defects
Results Gallery
Segmentation results showing Apple logo defect detection.
Training and evaluation demonstrations are omitted due to data confidentiality.
Demo Video
Apple logo inspection system demonstration.
Training and evaluation demonstrations are omitted due to data confidentiality.
End Result
- Reliable detection of logo-region defects with stable performance.
- Robust handling of reflective and low-contrast conditions.
- Production-ready system deployed in manufacturing.