Year: 2025
Type: article
Abstract: Improved YOLO11-based shaft-hole parts detection method In the assembly scene of shaft-hole parts, there are often occlusions and changes in the viewing angle, making real-time accurate detection of the target part there are certain difficulties, this paper is based on the structure of the YOLO11 network, respectively, through the introduction of the RepViT module, the edge information-based feature fusion module EFI and P-EfficientHead detection head, a multi-module fusion improved YOLO11 network for shaft hole assembly scenarios is proposed, and finally experimentally verified on the Pascal VOC dataset with, mAP50 and mAP50-95 are improved by 3.7% and 4%, and the precision and recall are improved by 2.9% and 2.5%, respectively, and the validation on the homemade dataset is also obtained with better results, and the final results show that the proposed multi-module fusion to improve YOLO11 network in this paper has better performance.
Authors: Yongpeng Tian, Bo Liu, Bingyuan Zhu, Jian Zhang
Institution: Tongji University
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4409772848