Skip to content

Add kernel splatter

John Biddiscombe requested to merge biddisco/vtk-m:add-gaussian-splatter into master

This replaces !135 (merged) with a more complete implementation and some kernel classes.

The algorithm used generates a list of all the voxels affected by the splat, then sorts and reduces them to produce the final splat contribution which is then applied into the final volume, it can be inefficient when many voxels are affected as it uses a considerable amount of memory.

An atomic implementation is underway.

Merge request reports