Skip to content
Remy Mellet edited this page Jan 21, 2021 · 28 revisions

As of version 2019.1.0 of meshroom, it is possible to do a reconstruction without using the Depthmap node and thus without the need to use a Nvidia graphic card (depthmap requires CUDA). It is much faster than depth map but the resulting mesh is low quality, so it is still recommended that the depthmap is used to generate the mesh if possible.

To set up the node sequence for draft meshing do the following:

  1. Delete the nodes PrepareDenseScene, Depthmap and DepthmapFilter
  2. Connect output from StructureFromMotion to input on Meshing
  3. Press the Auto-Layout button (located in the bottom left)

This is the resulting node sequence:

You should use the HIGH preset on the FeatureExtraction node to get enough density for the Meshing. See Reconstruction-parameters

Quality comparison:

Monstree dataset
Default reconstruction = with DepthMap
mesh

If you have only a few images, not selecting AKAZE in the SFM node can increase the point cloud density. Adjust Describer Preset in FeatureExtraction if necessary.