Please use this identifier to cite or link to this item:
Scopus Web of Science® Altmetric
Type: Conference paper
Title: A consensus-based framework for distributed Bundle Adjustment
Author: Eriksson, A.
Bastian, J.
Chin, T.
Isaksson, M.
Citation: Proceedings / CVPR, IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2016, vol.2016-December, pp.1754-1762
Publisher: IEEE
Issue Date: 2016
ISBN: 9781467388511
ISSN: 1063-6919
Conference Name: 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR 2016) (27 Jun 2016 - 30 Jun 2016 : Las Vegas, NV, USA)
Statement of
Anders Eriksson, John Bastian, Tat-Jun Chin and Mats Isaksson
Abstract: In this paper we study large-scale optimization problems in multi-view geometry, in particular the Bundle Adjustment problem. In its conventional formulation, the complexity of existing solvers scale poorly with problem size, hence this component of the Structure-from-Motion pipeline can quickly become a bottle-neck. Here we present a novel formulation for solving bundle adjustment in a truly distributed manner using consensus based optimization methods. Our algorithm is presented with a concise derivation based on proximal splitting, along with a theoretical proof of convergence and brief discussions on complexity and implementation. Experiments on a number of real image datasets convincingly demonstrates the potential of the proposed method by outperforming the conventional bundle adjustment formulation by orders of magnitude.
Keywords: Geometry, cameras, signal processing algorithms, computers, optimization, complexity theory, image reconstruction
Rights: © 2016 IEEE
DOI: 10.1109/CVPR.2016.194
Grant ID:
Published version:
Appears in Collections:Aurora harvest 3
Computer Science publications

Files in This Item:
File Description SizeFormat 
  Restricted Access
Restricted Access703.59 kBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.