Experimental Comparison of Two 6D Pose Estimation Algorithms in Robotic Fruit-Picking Tasks

被引:0
作者
Alterani, Alessio Benito [1 ]
Costanzo, Marco [1 ]
De Simone, Marco [1 ]
Federico, Sara [1 ]
Natale, Ciro [1 ]
机构
[1] Univ Campania Luigi Vanvitelli, Dipartimento Ingn, Via Roma 29, I-81031 Aversa, CE, Italy
关键词
pose estimation; robotic grasping; deep learning;
D O I
10.3390/robotics13090127
中图分类号
TP24 [机器人技术];
学科分类号
080202 ; 1405 ;
摘要
This paper presents an experimental comparison between two existing methods representative of two categories of 6D pose estimation algorithms nowadays commonly used in the robotics community. The first category includes purely deep learning methods, while the second one includes hybrid approaches combining learning pipelines and geometric reasoning. The hybrid method considered in this paper is a pipeline of an instance-level deep neural network based on RGB data only and a geometric pose refinement algorithm based on the availability of the depth map and the CAD model of the target object. Such a method can handle objects whose dimensions differ from those of the CAD. The pure learning method considered in this comparison is DenseFusion, a consolidated state-of-the-art pose estimation algorithm selected because it uses the same input data, namely, RGB image and depth map. The comparison is carried out by testing the success rate of fresh food pick-and-place operations. The fruit-picking scenario has been selected for the comparison because it is challenging due to the high variability of object instances in appearance and dimensions. The experiments carried out with apples and limes show that the hybrid method outperforms the pure learning one in terms of accuracy, thus allowing the pick-and-place operation of fruits with a higher success rate. An extensive discussion is also presented to help the robotics community select the category of 6D pose estimation algorithms most suitable to the specific application.
引用
收藏
页数:18
相关论文
共 37 条
  • [1] LEAST-SQUARES FITTING OF 2 3-D POINT SETS
    ARUN, KS
    HUANG, TS
    BLOSTEIN, SD
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1987, 9 (05) : 699 - 700
  • [2] A METHOD FOR REGISTRATION OF 3-D SHAPES
    BESL, PJ
    MCKAY, ND
    [J]. IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, 1992, 14 (02) : 239 - 256
  • [3] Modeling and slipping control of a planar slider
    Cavallo, Alberto
    Costanzo, Marco
    De Maria, Giuseppe
    Natale, Ciro
    [J]. AUTOMATICA, 2020, 115
  • [4] A Soft Gripper Design for Apple Harvesting with Force Feedback and Fruit Slip Detection
    Chen, Kaiwen
    Li, Tao
    Yan, Tongjie
    Xie, Feng
    Feng, Qingchun
    Zhu, Qingzhen
    Zhao, Chunjiang
    [J]. AGRICULTURE-BASEL, 2022, 12 (11):
  • [5] Costanzo M., 2023, P 2023 9 INT C CONTR, P901, DOI [10.1109/CoDIT58514.2023.10284072, DOI 10.1109/CODIT58514.2023.10284072]
  • [6] Non-Prehensile Manipulation Actions and Visual 6D Pose Estimation for Fruit Grasping Based on Tactile Sensing
    Costanzo, Marco
    De Simone, Marco
    Federico, Sara
    Natale, Ciro
    [J]. ROBOTICS, 2023, 12 (04)
  • [7] Control of robotic object pivoting based on tactile sensing
    Costanzo, Marco
    [J]. MECHATRONICS, 2021, 76
  • [8] Design and Calibration of a Force/Tactile Sensor for Dexterous Manipulation
    Costanzo, Marco
    De Maria, Giuseppe
    Natale, Ciro
    Pirozzi, Salvatore
    [J]. SENSORS, 2019, 19 (04)
  • [9] Towards Intelligent Fruit Picking with In-hand Sensing
    Dischinger, Lisa M.
    Cravetz, Miranda
    Dawes, Jacob
    Votzke, Callen
    VanAtter, Chelse
    Johnston, Matthew L.
    Grimm, Cindy M.
    Davidson, Joseph R.
    [J]. 2021 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2021, : 3285 - 3291
  • [10] Do TT, 2018, Arxiv, DOI [arXiv:1802.10367, 10.48550/ARXIV.1802.10367]