Design of a Library for Dense Matching

Matthieu Personnaz 1 Peter Sturm Frédéric Devernay
1 MOVI - Modeling, localization, recognition and interpretation in computer vision
GRAVIR - IMAG - Graphisme, Vision et Robotique, Inria Grenoble - Rhône-Alpes, CNRS - Centre National de la Recherche Scientifique : FR71
Abstract : The algorithms involved for a dense 3D reconstruction from two frames are highly dependent of the viewed scene. Furthermore, the user goals themselves induce should induce a specific choice. Indeed, the user focuses alternatively on the rendering, the running speed or the metrologic quality. Also, the aim of this paper is the design of a software library which allows to perform a dense matching between two images, by taking account both of the various user intents and world scenes. This leads first to the integration of several performant and possibly complex dense matching algorithms. The multiplicity and the complexity of the algorithms should however not be at the expense of the ergonomy of the software system. A futher goal is therefore to keep an easy and safe access to the library components. An iterative and object-oriented process is applied in order to cop with the design of a such library.
Document type :
Reports
Complete list of metadatas

https://hal.inria.fr/inria-00069900
Contributor : Rapport de Recherche Inria <>
Submitted on : Friday, May 19, 2006 - 6:29:22 PM
Last modification on : Friday, June 1, 2018 - 1:02:04 PM
Long-term archiving on : Saturday, April 3, 2010 - 9:42:46 PM

Identifiers

  • HAL Id : inria-00069900, version 1

Collections

Citation

Matthieu Personnaz, Peter Sturm, Frédéric Devernay. Design of a Library for Dense Matching. RT-0278, INRIA. 2003, pp.34. ⟨inria-00069900⟩

Share

Metrics

Record views

301

Files downloads

195