Fast and Accurate Multi-View Reconstruction by Multi-Stage Prioritized Matching

Markus Ylimäki, Juho Kannala, Jukka Holappa, Janne Heikkilä
Center for Machine Vision Research
University of Oulu

Sami S. Brandt
University of Copenhagen

The paper was published in the IET Computer Vision journal in 2015.


In this paper, we propose a multi-view stereo reconstruction method which creates a three-dimensional point cloud of a scene from multiple calibrated images captured from different viewpoints. The method is based on a prioritized match expansion technique, which starts from a sparse set of seed points, and iteratively expands them into neighboring areas by using multiple expansion stages. Each seed point represents a surface patch and has a position and a surface normal vector. The location and surface normal of the seeds are optimized using a homography-based local image alignment. The propagation of seeds is performed in a prioritized order in which the most promising seeds are expanded first and removed from the list of seeds. The first expansion stage proceeds until the list of seeds is empty. In the following expansion stages, the current reconstruction may be further expanded by finding new seeds near the boundaries of the current reconstruction. The prioritized expansion strategy allows efficient generation of accurate point clouds and our experiments show its benefits compared with nonprioritized expansion. In addition, a comparison to the widely used patch-based multi-view stereo software (PMVS) shows that our method is significantly faster and produces more accurate and complete reconstructions.


Software (zip)

Includes 32bit and 64bit Windows binaries with required dependencies. No support available.
When using the software in any publication please cite the paper in the title on this page.

Source codes (zip)

Source codes for the software. See the README.txt to get started.
When using the software in any publication please cite the paper in the title on this page.

Markus Ylimäki