ImagineMVS : 3D reconstruction in multiple view stereovision

This software was developed (or is under development) within the higher education and research community. Its stability can vary (see fields below) and its working state is not guaranteed.
Higher Edu - Research dev card
  • Creation or important update: 07/10/13
  • Minor correction: 07/10/13
  • Index card author: Pascal Monasse (LIGM)
  • Theme leader: Teresa Gomez-Diaz (LIGM)
General software features

A complete algorithmic pipeline for reconstruction of a 3D model from photographs.

The software takes as input calibrated views (that is, photographs with camera posed and oriented in space and internal parameters known). As output, it yields a textured 3D mesh. The steps are the following:

  • estimation of dense correspondences between images by plane sweeping,
  • triangulation to get a point cloud,
  • tetrahedral meshing of the point cloud and removal of hidden faces to get an initial mesh,
  • mesh optimization by incremental minimization of an energy balancing data fidelity and regularity,
  • creation from view fragments of a texture atlas to apply to the mesh.
Context in which the software is used

Internal use in IMAGINE project (École des Ponts ParisTech/CSTB).

This software is at the core of the start-up Acute3D.

Publications related to software

Vu Hoang Hiep, Renaud Keriven, Patrick Labatut and Jean-Philippe Pons. Towards high-resolution large-scale multi-view stereo, Proceedings of CVPR 2009, pp. 1430-1437.