Structure from motion using augmented Lagrangian robust factorization

Klaus Glashoff*, Michael M. Bronstein

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The classical Tomasi-Kanade method for Structure from Motion (SfM) based on measurement matrix factorization using SVD is known to perform poorly in the presence of occlusions and outliers. In this paper, we present an efficient approach by which we are able to deal with both problems at the same time. We use the Augmented Lagrangian alternative minimization method to solve iteratively a robust version of the matrix factorization approach. Experiments on synthetic and real data show the computational efficiency and good convergence of the method, which make it favorably compare to other approaches used in the SfM problem.

Original languageEnglish
Title of host publicationProceedings - 2nd Joint 3DIM/3DPVT Conference
Subtitle of host publication3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2012
Pages379-386
Number of pages8
DOIs
StatePublished - 2012
Externally publishedYes
Event2nd Joint 3DIM/3DPVT Conference: 3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2012 - Zurich, Switzerland
Duration: 13 Oct 201215 Oct 2012

Publication series

NameProceedings - 2nd Joint 3DIM/3DPVT Conference: 3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2012

Conference

Conference2nd Joint 3DIM/3DPVT Conference: 3D Imaging, Modeling, Processing, Visualization and Transmission, 3DIMPVT 2012
Country/TerritorySwitzerland
CityZurich
Period13/10/1215/10/12

Keywords

  • SfM
  • augmented Lagrangian
  • robust factorization
  • structure from motion

Fingerprint

Dive into the research topics of 'Structure from motion using augmented Lagrangian robust factorization'. Together they form a unique fingerprint.

Cite this