Projective Structure from Uncalibrated Images: Structure from Motion and Recognition

Amnon Shashua*

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

71 Scopus citations

Abstract

We address the problem of reconstructing 3-D space in a projective framework from two or more views, and the problem of artificially generating novel views of the scene from two given views (reprojection). We describe an invariance relation that provides a new descrijtlion of structure, which we call projective dejtlh, that is cajtlured by a single equation relating image point correspondences across two or more views and the homographies of two arbitrary virtual planes. The framework is based on knowledge of correspondence of features across views, is linear and extremely simple, and the computations of structure readily extend to overdetermination using multiple views. Experimental results demonstrate a high degree of accuracy in both tasks: Reconstruction and reprojection.

Original languageAmerican English
Pages (from-to)778-790
Number of pages13
JournalIEEE Transactions on Pattern Analysis and Machine Intelligence
Volume16
Issue number8
DOIs
StatePublished - Aug 1994

Bibliographical note

Funding Information:
Manuscript received December 4, 1992; revised February 18, 1994. This work was supported in part by the National Science Foundation (NSF) under Grant IRI-8900267. Recommended for acceptance by Associate Editor Y. Aloimonos.

Keywords

  • 3-D reconstruction from 2-D views
  • Visual recognition
  • algebraic and geometric invariants
  • projective geometry

Fingerprint

Dive into the research topics of 'Projective Structure from Uncalibrated Images: Structure from Motion and Recognition'. Together they form a unique fingerprint.

Cite this