Robotic Grasp Pose Detection Method Based on Multiscale Features

被引:2
|
作者
Wang, Zheng [1 ]
Leng, Longlong [2 ]
Zhou, Xianming [3 ]
Zhao, Yanwei [2 ]
机构
[1] Hangzhou City Univ, Sch Comp & Computat Sci, 51 Huzhou St, Hangzhou, Zhejiang, Peoples R China
[2] Hangzhou City Univ, Dept Mech Engn, 51 Huzhou St, Hangzhou, Zhejiang, Peoples R China
[3] Zhejiang Univ Technol, Coll Mech Engn, 255 Liuhe Rd, Hangzhou, Zhejiang, Peoples R China
基金
中国国家自然科学基金;
关键词
Deep learning; grasp detection; multiscale feature; unstructured environment; YOLOv3;
D O I
10.1142/S021984362350010X
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
A robotic grasp detection algorithm based on multiscale features is proposed for autonomous robotic grasping in an unstructured environment. The grasp detection model borrowed the YOLOv3 object detection algorithm and retained the original idea of multiscale detection to improve the perception ability of the grasp rectangle on different scales. Squeeze and excitation blocks were embedded into the Residual Networks (ResNet) structure of the original model, with deformable convolution (DC) introduced, so that the model attained stronger feature extraction ability to cope with more complex grasp detection tasks. Meanwhile, the prediction of the direction angle was transformed into a combination of classification and regression, achieving the prediction of the direction angle of the grabbing frame under different postures. The model was simulated on the Cornell grasp dataset. The results demonstrate that the algorithm in this study can effectively balance the accuracy and efficiency of detection and can migrate the prediction of the grasp rectangle to new objects. The results of online grasp experiments on a Baxter robot show that the average grasp success rate of 93% is achieved for 10 different objects, demonstrating the practical feasibility of the algorithm.
引用
收藏
页数:24
相关论文
共 50 条
  • [1] Robotic Grasp Detection Using Structure Prior Attention and Multiscale Features
    Chen, Lu
    Niu, Mingdi
    Yang, Jing
    Qian, Yuhua
    Li, Zhuomao
    Wang, Keqi
    Yan, Tao
    Huang, Panfeng
    [J]. IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, 2024,
  • [2] Grasp generation method based on multiscale features fusion and grasp quality assessment
    Gao, Xiang
    Xie, Haisheng
    Zhu, Bo
    Xu, Guozheng
    [J]. Yi Qi Yi Biao Xue Bao/Chinese Journal of Scientific Instrument, 2023, 44 (07): : 101 - 111
  • [3] Robotic Grasp Pose Detection Using Deep Learning
    Caldera, Shehan
    Rassau, Alexander
    Chai, Douglas
    [J]. 2018 15TH INTERNATIONAL CONFERENCE ON CONTROL, AUTOMATION, ROBOTICS AND VISION (ICARCV), 2018, : 1966 - 1972
  • [4] MTGrasp: Multiscale 6-DoF Robotic Grasp Detection
    Yu, Sheng
    Zhai, Di-Hua
    Xia, Yuanqing
    [J]. IEEE-ASME TRANSACTIONS ON MECHATRONICS, 2024,
  • [5] Grasp Pose Detection Based on Shape Simplification
    Cao, Chuqing
    Liu, Hanwei
    [J]. INTERNATIONAL JOURNAL OF HUMANOID ROBOTICS, 2021, 18 (03)
  • [6] Robotic Grasp Detection With 6-D Pose Estimation Based on Graph Convolution and Refinement
    Yu, Sheng
    Zhai, Di-Hua
    Xia, Yuanqing
    Wang, Wei
    Zhang, Chengyu
    Zhao, Shiqi
    [J]. IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS, 2024, 54 (06): : 3783 - 3795
  • [7] Robotic Grasp Detection Based on Transformer
    Dong, Mingshuai
    Bai, Yuxuan
    Wei, Shimin
    Yu, Xiuli
    [J]. INTELLIGENT ROBOTICS AND APPLICATIONS (ICIRA 2022), PT IV, 2022, 13458 : 437 - 448
  • [8] Detection Method of Manipulator Grasp Pose Based on RGB-D Image
    Huang, Cheng
    Pang, Zhen
    Xu, Jiazhong
    [J]. NEURAL PROCESSING LETTERS, 2024, 56 (04)
  • [9] A novel robotic grasp detection method based on region proposal networks
    Song, Yanan
    Gao, Liang
    Li, Xinyu
    Shen, Weiming
    [J]. ROBOTICS AND COMPUTER-INTEGRATED MANUFACTURING, 2020, 65
  • [10] Generative model based robotic grasp pose prediction with limited dataset
    Priya Shukla
    Nilotpal Pramanik
    Deepesh Mehta
    G. C. Nandi
    [J]. Applied Intelligence, 2022, 52 : 9952 - 9966