Research on a Fusion Technique of YOLOv8-URE-Based 2D Vision and Point Cloud for Robotic Grasping in Stacked Scenarios
Source abstract
Abstract
In industrial robotic grasping tasks, traditional 3D point cloud registration and pose estimation methods often struggle with low efficiency and limited accuracy in stacked and cluttered environments. To address these challenges, this paper proposes a grasp pose estimation algorithm that integrates 2D object detection based on YOLOv8-URE with 3D point cloud registration. In the detection stage, the method enhances object feature perception and localization by optimizing the receptive field structure and introducing attention mechanisms. It also employs an efficient multi-scale feature fusion strategy to improve bounding box regression accuracy. During point cloud processing, target centers predicted by the detector guide rapid segmentation, followed by robust registration techniques to estimate precise object poses. Experimental results demonstrate that YOLOv8-URE improves detection accuracy by 9.21% compared to YOLOv8n, reduces registration time by 60.5%, and significantly increases grasp success rates, proving its reliability and effectiveness in industrial scenarios.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Robot Manipulation and Learning · Robotics and Sensor-Based Localization · Robotics and Automated Systems
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
School of Mechanical Engineering, Hubei University of Technology, Wuhan 430068, China
Hubei Key Laboratory of Modern Manufacturing Quality Engineering, Wuhan 430068, China
School of Mechanical and Electrical Engineering, Wuhan Donghu University, Wuhan 430212, China
Funding
Natural Science Foundation of Hubei Province (Youth Program) · Hubei Agricultural Machinery Equipment Shortcomings Tackling Project “Research, Development, and Promotion of Key Technologies and Equipment for Aquatic Product Processing” · Youth Talent Project of the Science and Technology Research Program of the Hubei Provincial Department of Education
References (21)
Show source-provided references
1. Application and industrial development of machine vision in intelligent manufacturing
2. Research on control of intelligent logistics sorting robot based on laser vision guidance
3. Visual servo control of agricultural robot parallel picking arm
4. Lu, Z. (2020). Research on stacking object grasping method based on deep learning. [Master’s Thesis, Guangdong University of Technology]. (In Chinese).
5. Optimal grasp posture detection method for robots based on deep learning10.1007/s11430-018-9387-2 ↗
6. Guo, D., Kong, T., Sun, F., and Liu, H. (2016, January 16–21). Object discovery and grasp detection with a shared convolutional neural network. Proceedings of the 2016 IEEE International Conference on Robotics and Automation, Stockholm, Sweden.
7. Geng, Z., and Chen, G. (2022, January 27–29). A novel real-time grasping method combined with YOLO and GDFCN. Proceedings of the 2022 IEEE 10th Joint International Information Technology and Artificial Intelligence Conference (ITAIC), Beijing, China.10.1109/ITAIC54216.2022.9836458 ↗
8. Lightweight model-based 6D pose estimation of drones
9. Pose estimation for workpieces in complex stacking industrial scene based on RGB images
10. HRPose: Real-time high-resolution 6D pose estimation network using knowledge distillation10.23919/cje.2021.00.211 ↗
11. Wang, G., Manhardt, F., Tombari, F., and Ji, X. (2021, January 19–25). GDR-Net: Geometry-guided direct regression network for monocular 6D object pose estimation. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.10.1109/CVPR46437.2021.01634 ↗
12. Xu, W., and Wan, Y. (2024). ELA: Efficient local attention for deep convolutional neural networks. arXiv.
13. Zhang, H., and Zhang, S. (2023). Shape-Iou: More accurate metric considering bounding box shape and scale. arXiv.
14. Ding, X., Zhang, Y., Ge, Y., Zhao, S., Song, L., Yue, X., and Shan, Y. (2024, January 17–21). UniRepLKNet: A universal perception large-kernel convNet for audio, video, point cloud, time-series, and image recognition. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA.10.1109/CVPR52733.2024.00527 ↗
15. Ding, X., Zhang, X., Ma, N., Han, J., Ding, G., and Sun, J. (2021, January 19–25). RepVGG: Making VGG-style ConvNets great again. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA.10.1109/CVPR46437.2021.01352 ↗
16. Ioffe, S., and Szegedy, C. (2015, January 6–11). Batch normalization: Accelerating deep network training by reducing internal covariate shift. Proceedings of the 32nd International Conference on Machine Learning (ICML), Lille, France.
17. Ding, X., Zhang, X., Han, J., and Ding, G. (2022, January 19–24). Scaling up your kernels to 31×31: Revisiting large kernel design in CNNs. Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, LA, USA.10.1109/CVPR52688.2022.01166 ↗
18. Draelos, R.L., and Carin, L. (2020). Use HiResCAM instead of Grad-CAM for faithful explanations of convolutional neural networks. arXiv.
19. Xu, X., Jiang, Y., Chen, W., Huang, Y., Zhang, Y., and Sun, X. (2022). Damo-YOLO: A report on real-time object detection design. arXiv.
20. Liu, W., Lu, H., Fu, H., and Cao, Z. (2023, January 2–7). Learning to upsample by learning to sample. Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV), Paris, France.10.1109/ICCV51070.2023.00554 ↗
21. Baranchuk, D., Babenko, A., and Malkov, Y. (2018, January 8–14). Revisiting the inverted indices for billion-scale approximate nearest neighbors. Proceedings of the European Conference on Computer Vision (ECCV), Munich, Germany.10.1007/978-3-030-01258-8_13 ↗