DOI QR코드

DOI QR Code

Enhancing 3D Excavator Pose Estimation through Realism-Centric Image Synthetization and Labeling Technique

  • Tianyu Liang (Department of Civil & Mineral Engineering, University of Toronto) ;
  • Hongyang Zhao (Department of Civil & Mineral Engineering, University of Toronto) ;
  • Seyedeh Fatemeh Saffari (Department of Civil & Mineral Engineering, University of Toronto) ;
  • Daeho Kim (Department of Civil & Mineral Engineering, University of Toronto)
  • Published : 2024.07.29

Abstract

Previous approaches to 3D excavator pose estimation via synthetic data training utilized a single virtual excavator model, low polygon objects, relatively poor textures, and few background objects, which led to reduced accuracy when the resulting models were tested on differing excavator types and more complex backgrounds. To address these limitations, the authors present a realism-centric synthetization and labeling approach that synthesizes results with improved image quality, more detailed excavator models, additional excavator types, and complex background conditions. Additionally, the data generated includes dense pose labels and depth maps for the excavator models. Utilizing the realism-centric generation method, the authors achieved significantly greater image detail, excavator variety, and background complexity for potentially improved labeling accuracy. The dense pose labels, featuring fifty points instead of the conventional four to six, could allow inferences to be made from unclear excavator pose estimates. The synthesized depth maps could be utilized in a variety of DNN applications, including multi-modal data integration and object detection. Our next step involves training and testing DNN models that would quantify the degree of accuracy enhancement achieved by increased image quality, excavator diversity, and background complexity, helping lay the groundwork for broader application of synthetic models in construction robotics and automated project management.

Keywords

References

  1. OSHA (2013), "https://www.osha.gov/SLTC/etools/construction/struckby/mainpage. html," OSHA Construction eTools.
  2. C.-J. Liang, K. M. Lundeen, W. Mcgee, C. C. Menassa, S. Lee, and V. R. Kamat, "A Vision-based Marker-less Pose Estimation System 1 for Articulated Construction Robots," 2019.
  3. H. Luo, M. Wang, P. K. Y. Wong, and J. C. P. Cheng, "Full body pose estimation of construction equipment using computer vision and deep learning techniques," Autom Constr, vol. 110, Feb. 2020, doi: 10.1016/j.autcon.2019.103016.
  4. A. Assadzadeh, M. Arashpour, I. Brilakis, T. Ngo, and E. Konstantinou, "Vision-based excavator pose estimation using synthetically generated datasets with domain randomization," Autom Constr, vol. 134, Feb. 2022, doi: 10.1016/j.autcon.2021.104089.
  5. A. Assadzadeh, M. Arashpour, H. Li, R. Hosseini, F. Elghaish, and S. Baduge, "Excavator 3D pose estimation using deep learning and hybrid datasets," Advanced Engineering Informatics, vol. 55, p. 101875, Jan. 2023, doi: 10.1016/j.aei.2023.101875.
  6. B. Mahmood, S. U. Han, and J. Seo, "Implementation experiments on convolutional neural network training using synthetic images for 3D pose estimation of an excavator on real images," Autom Constr, vol. 133, Jan. 2022, doi: 10.1016/j.autcon.2021.103996.