Mask Detection Method and Apparatus
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-04-24
- Publication Date
- 2026-05-26
AI Technical Summary
Existing technologies cannot meet the high-precision, high-efficiency, and fully automated mask defect detection requirements of OLED production lines. Manual visual inspection is inefficient and prone to missed or false detections, while traditional machine vision inspection is insufficient in detecting minute defects and cannot be adapted to different models and mesh parameters.
A mask detection method is adopted. By constructing a defect dataset, feature extraction and pyramid network fusion are performed to build a defect detection model. The model is optimized by combining Focal Loss, DIoU Loss and Dice Loss to achieve automated detection. Based on the detection results, process adjustment instructions are generated and fed back to the processing equipment.
It enables automated detection of mask defects, reduces personnel workload, improves production efficiency, corrects process deviations in a timely manner, avoids the generation of batch defective products, and reduces production costs.
Smart Images

Figure CN122089738A_ABST
Abstract
Citation Information
Patent Citations
Netting device and netting method for mask template
CN106086784A
Net expansion equipment
CN107732012A
Deep learning-based tomato disease and insect pest detection and identification method
CN116630803A
Mask plate defect detection method, system and equipment based on deep learning and medium
CN120807506A
Greenhouse tomato fruit maturity instance segmentation method based on dual-backbone network fusion
CN121789210A