Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

3D Reconstruction in Noisy Agricultural Environments: A Bayesian Optimization Perspective for View Planning

Athanasios Bacharis, Konstantinos D. Polyzos, Henry J. Nelson, Georgios B. Giannakis, Nikolaos Papanikolopoulos

TL;DR

This work puts forth an adaptive Bayesian optimization algorithm for accurate 3D reconstruction in the presence of noise, that accounts for the existing noise of the environment, without requiring its closed-form expression.

Abstract

3D reconstruction is a fundamental task in robotics that gained attention due to its major impact in a wide variety of practical settings, including agriculture, underwater, and urban environments. This task can be carried out via view planning (VP), which aims to optimally place a certain number of cameras in positions that maximize the visual information, improving the resulting 3D reconstruction. Nonetheless, in most real-world settings, existing environmental noise can significantly affect the performance of 3D reconstruction. To that end, this work advocates a novel geometric-based reconstruction quality function for VP, that accounts for the existing noise of the environment, without requiring its closed-form expression. With no analytic expression of the objective function, this work puts forth an adaptive Bayesian optimization algorithm for accurate 3D reconstruction in the presence of noise. Numerical tests on noisy agricultural environments showcase the merits of the proposed approach for 3D reconstruction with even a small number of available cameras.

3D Reconstruction in Noisy Agricultural Environments: A Bayesian Optimization Perspective for View Planning

TL;DR

This work puts forth an adaptive Bayesian optimization algorithm for accurate 3D reconstruction in the presence of noise, that accounts for the existing noise of the environment, without requiring its closed-form expression.

Abstract

3D reconstruction is a fundamental task in robotics that gained attention due to its major impact in a wide variety of practical settings, including agriculture, underwater, and urban environments. This task can be carried out via view planning (VP), which aims to optimally place a certain number of cameras in positions that maximize the visual information, improving the resulting 3D reconstruction. Nonetheless, in most real-world settings, existing environmental noise can significantly affect the performance of 3D reconstruction. To that end, this work advocates a novel geometric-based reconstruction quality function for VP, that accounts for the existing noise of the environment, without requiring its closed-form expression. With no analytic expression of the objective function, this work puts forth an adaptive Bayesian optimization algorithm for accurate 3D reconstruction in the presence of noise. Numerical tests on noisy agricultural environments showcase the merits of the proposed approach for 3D reconstruction with even a small number of available cameras.
Paper Structure (13 sections, 16 equations, 6 figures, 1 table)

This paper contains 13 sections, 16 equations, 6 figures, 1 table.

Table of Contents

  1. Introduction
  2. Related Work
  3. View Planning
  4. Bayesian Optimization
  5. Problem Formulation
  6. Optimization Method
  7. BO with single GP and EI
  8. BO with ensemble GPs and adaptive EI
  9. Experiments
  10. Implementation Details
  11. Evaluation
  12. Conclusions
  13. Acknowledgements

Figures (6)

  • Figure 1: Example of a noisy environment with four drones aiming to carry out the 3D reconstruction task (drone image ).
  • Figure 2: EGP-VP in a nutshell.
  • Figure 3: Example of the three environmental cases.
  • Figure 4: Simple regret for (a) 1-plant, (b) 3-plant, and (c) 6-plant scenarios.
  • Figure 5: Reconstruction results from all competing methods shown on the 3-plant case from the same viewpoint to demonstrate differences.
  • ...and 1 more figures