All Title Author
Keywords Abstract

Publish in OALib Journal
ISSN: 2333-9721
APC: Only $99

ViewsDownloads

Relative Articles

More...

基于多物理特征的抓取检测方法
Grasping Detection Method Based on Multiple Physical Features

DOI: 10.12677/dsc.2024.132005, PP. 45-53

Keywords: 点云,位姿估计,多物理特征,深度学习
Point Cloud
, Pose Estimation, Multiple Physical Features, Deep Learning

Full-Text   Cite this paper   Add to My Lib

Abstract:

为解决在无约束、非结构化的多物体场景下对物体生成可靠抓取姿态的问题,提高机械臂在多场景下抓取检测的准确率,我们提出了使用场景的多物理特征进行抓取检测的评估方式,基于PointNet 网络提出了一种前置–后置联合网络抓取姿态估计算法,该算法利用多物理特征从目标点云中生成评估得到的置信度分数对场景进行进一步抓取检测得到可靠的抓取姿态。由于该抓取检测方法比以往的抓取检测评估更加细致,从而可以更好地从场景中得到精确的抓取姿态。实验结果表明:该算法在复杂的多物体场景中,能够生成有效的抓取姿态,且较同类型算法抓取成功率有所提升,能够应用于工业机器人的抓取任务。
In order to solve the problem of generating a reliable grasping attitude for objects in unconstrained and unstructured multi-object scenes and improve the accuracy of grasping detection of mechanical arm in multiple scenarios, we put forward an evaluation method of grasping evaluation method using multiple physical features of scenes. Based on PointNet network, we put forward a front-post joint network grasping attitude estimation algorithm. The algorithm uses multiple physical features to generate the confidence score obtained from the target point cloud to further capture and detect the scene to get a reliable grasping attitude. Because this grasping detection method is more detailed than the previous grasping detection evaluation, it can get an accurate grasp posture from the scene. The experimental results show that the algorithm can generate an effective grasping attitude in complex multi-object scenes, and the grasping success rate is improved compared with the same algorithm, which can be applied to the grasping task of industrial robots.

References

[1]  Mousavian, A., Eppner, C. and Fox, D. (2019) 6-DOF GraspNet: Variational Grasp Generation for Object Manipulation. 2019 IEEE/CVF International Conference on Computer Vision (ICCV), Seoul, 27 October-2 November 2019, 2901-2910.
https://doi.org/10.1109/ICCV.2019.00299
[2]  Sundermeyer, M., Mousavian, A., Triebel, R. and Fox, D. (2021) Contact-GraspNet: Efficient 6-DoF Grasp Generation in Cluttered Scenes. 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi’an, 30 May-5 June 2021, 13438-13444.
https://doi.org/10.1109/ICRA48506.2021.9561877
[3]  Wu, W., Qi, Z. and Li, F. (2020) PointConv: Deep Convolutional Networks on 3D Point Clouds. 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Long Beach, CA, 15-20 June 2019, 9613-9622.
https://doi.org/10.1109/CVPR.2019.00985
[4]  Morrison, D., Corke, P. and Leitner, J. (2018) Closing the Loop for Robotic Grasping: A Real-Time, Generative Grasp Synthesis Approach. arXiv: 1804.05172.
https://doi.org/10.15607/RSS.2018.XIV.021
[5]  Liang, H., Ma, X., Li, S., G?rner, M., Tang, S., Fang, B., Sun, F. and Zhang, J. (2019) PointNetGPD: Detecting Grasp Configurations from Point Sets. 2019 International Conference on Robotics and Automation (ICRA), Montreal, QC, 20-24 May 2019, 3629-3635.
https://doi.org/10.1109/ICRA.2019.8794435
[6]  Yu, S., Zhai, D.-H. and Xia, Y. (2023) CGNet: Robotic Grasp Detection in Heavily Cluttered Scenes. IEEE/ASME Transactions on Mechatronics, 28, 884-894.
https://doi.org/10.1109/TMECH.2022.3209488
[7]  瞿孝昌. 基于点云特征的工业机械臂六自由度抓取方法研究[D]: [硕士学位论文]. 徐州: 中国矿业大学, 2022.
[8]  Wang, C., Fang, H.-S., Gou, M., Fang, H., Gao, J. and Lu, C. (2021) Graspness Discovery in Clutters for Fast and Accurate Grasp Detection. 2021 IEEE/CVF International Conference on Computer Vision (ICCV), Montreal, QC, 10-17 October 2021, 15944-15953.
https://doi.org/10.1109/ICCV48922.2021.01566
[9]  阮国强, 曹雏清. 基于PointNet 的机器人抓取姿态估计[J]. 仪表技术与传感器, 2023(5): 44-48.
[10]  Fang, H.-S., Wang, C., Gou, M. and Lu, C. (2020) GraspNet-1Billion: A Large-Scale Benchmark for General Object Grasping. 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, 13-19 June 2020, 11441-11450.
https://doi.org/10.1109/CVPR42600.2020.01146
[11]  Qi, C.R., Yi, L., Su, H. and Guibas, L.J. (2017) PointNet : Deep Hierarchical Feature Learning on Point Sets in a Metric Space. arXiv: 1706.02413.
[12]  楚红雨, 冷齐齐, 张晓强, 等. 融入注意力机制的多模特征机械臂抓取位姿检测[J]. 控制与决策2024, 39(3): 777-785.
https://doi.org/10.13195/j.kzyjc.2022.0812
[13]  Zhao, B., Zhang, H., Lan, X., Wang, H., Tian, Z. and Zheng, N. (2021) REGNet: REgion-Based Grasp Network for End-to-End Grasp Detection in Point Clouds. 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi’an, 30 May-5 June 2021, 13474-13480.
https://doi.org/10.1109/ICRA48506.2021.9561920
[14]  Lu, Y., Deng, B., Wang, Z., Zhi, P., Li, Y. and Wang, S. (2022) Hybrid Physical Metric For 6-DoF Grasp Pose Detection. 2022 International Conference on Robotics and Automation (ICRA), Philadelphia, PA, 23-27 May 2022, 8238-8244.
https://doi.org/10.1109/ICRA46639.2022.9811961
[15]  高翔, 谢海晟, 朱博, 徐国政. 基于多尺度特征融合和抓取质量评估的抓取生成方法[J]. 仪器仪表学报, 2023, 44(7): 101-111.
[16]  Shi, C., Miao, C., Zhong, X., Zhong, X., Hu, H. and Liu, Q. (2022) Pixel-Reasoning-Based Robotics Fine Grasping for Novel Objects with Deep EDINet Structure. Sensors, 22, Article 4283.
https://doi.org/10.3390/s22114283

Full-Text

comments powered by Disqus

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133

WeChat 1538708413