一种基于深度学习的刨花板表面缺陷检测方法
By improving the YOLOv11 target detection model, utilizing an efficient multi-scale attention mechanism and a weighted feature pyramid network, combined with the Shape-NWD loss function, the problem of low accuracy and efficiency in particleboard surface defect detection was solved, achieving high-precision detection of various types and minute defects, and reducing labor costs.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING FORESTRY UNIV
- Filing Date
- 2025-02-25
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies for detecting defects on particleboard surfaces suffer from low accuracy, low efficiency, and high labor costs, and are particularly inadequate for handling a wide variety of complex and minute defects.
An improved YOLOv11 target detection model is adopted. By constructing an efficient multi-scale attention mechanism (EMA) and a weighted feature pyramid network (BiFPN), combined with the Shape-NWD loss function, the feature extraction and fusion capabilities are improved, thereby enhancing the detection accuracy of complex and small defects.
It improves the accuracy and efficiency of particleboard surface defect detection, reduces labor costs, can accurately detect various types and minute defects, and enhances adaptability to complex scenarios.
Smart Images
Figure CN120147248B_ABST