A weld seam recognition and positioning method and system based on YOLOv11
By improving the YOLOv11 network architecture and adopting Starnet and FreqFusion modules, efficient and accurate weld seam identification and positioning are achieved, solving the identification problem on resource-constrained devices and improving multi-task synchronous prediction capabilities.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HEBEI UNIVERSITY
- Filing Date
- 2026-03-03
- Publication Date
- 2026-07-17
AI Technical Summary
Existing weld seam recognition technologies are difficult to deploy efficiently on resource-constrained edge devices. Furthermore, traditional methods are sensitive to imaging noise, resulting in decreased recognition accuracy and insufficient generalization ability, making them difficult to adapt to various types of weld seams and complex working conditions.
Based on the YOLOv11-pose network architecture, the backbone network is replaced with the Starnet structure, and the FreqFusion feature fusion module is introduced into the neck network. Combined with the key point prediction branch and depthwise separable convolution, multi-task synchronous prediction is achieved.
While reducing computational complexity and the number of parameters, it improves the accuracy and efficiency of weld seam identification and positioning, making it suitable for deployment on edge devices with limited computing resources.
Smart Images
Figure CN122415422A_ABST