Shuenhoy/vtk-m
1
0
Fork 0
mirror of https://gitlab.kitware.com/vtk/vtk-m synced 2024-07-11 09:33:37 +00:00
Code Issues 2 Actions Packages Projects Releases Wiki Activity

Fix divide by zero in CanvasRayTracer

Browse Source 
The divide by zero happened when mapping a surface to a depth buffer.
Some rays terminated at the origin, which is a singularity in the
project matrix.

This might be indicative of a larger problem. Rays really shouldn't
terminate before the near plane.
This commit is contained in:
Kenneth Moreland 2021-07-12 09:25:07 -06:00
parent 83369ed95a
commit 4c1514bb3a
1 changed files with 18 additions and 8 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

26
vtkm/rendering/CanvasRayTracer.cxx
View File

@ -59,15 +59,25 @@ public:
point[2] = static_cast<vtkm::Float32>(intersection[2]);
point[3] = 1.f;
vtkm::Vec4f_32 newpoint;
newpoint = vtkm::MatrixMultiply(this->ViewProjMat, point);
newpoint[0] = newpoint[0] / newpoint[3];
newpoint[1] = newpoint[1] / newpoint[3];
newpoint[2] = newpoint[2] / newpoint[3];
vtkm::Float32 depth;
{
vtkm::Vec4f_32 newpoint;
newpoint = vtkm::MatrixMultiply(this->ViewProjMat, point);
if (newpoint[3] > 0)
{
depth = 0.5f * (newpoint[2] / newpoint[3]) + 0.5f;
}
else
{
// This condition can happen when the ray is at the origin (inDepth = 0), which is a
// singularity in the projection matrix. I'm not sure this is the right think to do since
// it looks like depth is supposed to be between 0 and 1. It seems wrong that you would
// ever get a ray in front of the near plane, so the "right" solution may be to fix this
// elsewhere.
depth = vtkm::NegativeInfinity32();
}
}
vtkm::Float32 depth = newpoint[2];
depth = 0.5f * (depth) + 0.5f;
vtkm::Vec4f_32 color;
color[0] = static_cast<vtkm::Float32>(colorBufferIn.Get(index * 4 + 0));
color[1] = static_cast<vtkm::Float32>(colorBufferIn.Get(index * 4 + 1));