Estimation method of category-level multi-object rigid body 6D pose

被引:0
作者
Cheng, Shuo [1 ]
Jia, Di [1 ,2 ]
Yang, Liu [1 ]
He, Dekun
机构
[1] Liaoning Tech Univ, Sch Elect & Informat Engn, Huludao 125105, Peoples R China
[2] Liaoning Tech Univ, Ordos Inst, Ordos 017000, Peoples R China
基金
中国国家自然科学基金;
关键词
6D pose estimation; multi-objective single-stage network; multi-drop feature extraction layer; feature selection; composite data;
D O I
10.37188/CJLCD.2024-0182
中图分类号
O7 [晶体学];
学科分类号
0702 ; 070205 ; 0703 ; 080501 ;
摘要
In order to solve the problems of poor scalability, low generality and high computational cost of the traditional method using single object CNN model, and optimize the performance of multi-objective method. In this paper, a single-stage network architecture for multi-objective 6D attitude estimation is proposed, and a multi-branch feature extraction decoder is designed to capture and aggregate detailed features effectively. This paper proposes a feature optimization and screening module, which filters input features to extract multi-scale features. Combining the above two, a new feature pyramid structure is designed to improve the overall performance of the network and improve the pose estimation effect of occlusion. The experiments are carried out on synthetic data set LINEMOD and Occluded LINEMOD. The results show that the proposed method has achieved significant improvement in the processing of blocked object scenes. Compared with the most advanced methods such as PyraPose, SD-Pose and CASAPose, the proposed method has increased the ADD/S-Recall index by 43. 1 degrees o, 16. 1 degrees o and 12 degrees o, respectively. It performed better when the number of targets is small, increasing performance by 17 degrees o when the number of targets is 4. The ablation experiment further verifies the effectiveness of each module. By introducing multi-branch feature extraction decoder, feature optimization and screening module, and feature pyramid structure, the proposed single- stage multi-objective network architecture can process any number of targets by training only one network, and can perform 6D pose estimation better under the condition of synthetic data. Experimental results verify the effectiveness of the proposed method.
引用
收藏
页码:457 / 471
页数:15
相关论文
共 24 条
  • [1] InstancePose: Fast 6DoF Pose Estimation for Multiple Objects from a Single RGB Image
    Aing, Lee
    Lie, Wen-Nung
    Chiang, Jui-Chiu
    Lin, Guo-Shiang
    [J]. 2021 IEEE/CVF INTERNATIONAL CONFERENCE ON COMPUTER VISION WORKSHOPS (ICCVW 2021), 2021, : 2621 - 2630
  • [2] Brachmann E, 2014, LECT NOTES COMPUT SC, V8690, P536, DOI 10.1007/978-3-319-10605-2_35
  • [3] ConvPoseCNN: Dense Convolutional 6D Object Pose Estimation
    Capellen, Catherine
    Schwarz, Max
    Behnke, Sven
    [J]. PROCEEDINGS OF THE 15TH INTERNATIONAL JOINT CONFERENCE ON COMPUTER VISION, IMAGING AND COMPUTER GRAPHICS THEORY AND APPLICATIONS, VOL 5: VISAPP, 2020, : 162 - 172
  • [4] PoseRBPF: A Rao-Blackwellized Particle Filter for 6-D Object Pose Tracking
    Deng, Xinke
    Mousavian, Arsalan
    Xiang, Yu
    Xia, Fei
    Bretl, Timothy
    Fox, Dieter
    [J]. IEEE TRANSACTIONS ON ROBOTICS, 2021, 37 (05) : 1328 - 1342
  • [5] DUMOULIN V, 2017, P 5 INT C LEARN REPR
  • [6] GARD N, 2022, P 33 BRIT MACH VIS C, P899
  • [7] Gu Wang, 2020, Computer Vision - ECCV 2020. 16th European Conference. Proceedings. Lecture Notes in Computer Science (LNCS 12346), P108, DOI 10.1007/978-3-030-58452-8_7
  • [8] Hinterstoisser S, 2012, LECT NOTES COMPUT SC, V7585, P593, DOI 10.1007/978-3-642-33885-4_60
  • [9] HOU T B, 2020, arXiv
  • [10] Segmentation-driven 6D Object Pose Estimation
    Hu, Yinlin
    Hugonot, Joachim
    Fua, Pascal
    Salzmann, Mathieu
    [J]. 2019 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR 2019), 2019, : 3380 - 3389