Next Best View Planning via Reinforcement Learning for Scanning of Arbitrary 3D Shapes

S. G. Potapova, A. V. Artemov, S. V. Sviridov, D. A. Musatkina, D. N. Zorin, E. V. Burnaev

Research output: Contribution to journalArticlepeer-review

Abstract

Abstract: Reconstructing 3D objects from scanned measurements is a fundamental task in computer vision. A central factor for the effectiveness of 3D reconstruction is the selection of sensor views for scanning. The latter remains an open problem in the 3D geometry processing area, known as the next-best-view planning problem, and is commonly approached by combinatorial or greedy methods. In this work, we propose a reinforcement learning-based approach to sequential next-best-view planning. The method is implemented based on the gym environment including 3D reconstruction, next-best-scan planning, and image acquisition features. We demonstrate this method to outperform the baselines in terms of the number of required scans and the obtained 3D mesh reconstruction accuracy.

Original languageEnglish (US)
Pages (from-to)1484-1490
Number of pages7
JournalJournal of Communications Technology and Electronics
Volume65
Issue number12
DOIs
StatePublished - Dec 2020

Keywords

  • 3D model
  • CAD model
  • depth map
  • mesh
  • next best view
  • reinforcement learning

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Radiation
  • Condensed Matter Physics
  • Electrical and Electronic Engineering

Fingerprint Dive into the research topics of 'Next Best View Planning via Reinforcement Learning for Scanning of Arbitrary 3D Shapes'. Together they form a unique fingerprint.

Cite this