Transperant isosurface rendering - Moving from pyvista to pure VTK

Support
1

Hey guys, I’m just having some difficulties moving over from pyvista to pure VTK (so that I can eventually move into c++). What I’ve been attempting is rendering a grid of points with associated scalar values as a series of transparent isosurfaces. Illustrated below.

image
image858×684 53.1 KB

And I’ve managed to do this in Pyvista with just a few LOC

X, Y, Z = np.meshgrid(x0range, y0range, z0range, indexing='ij')
mesh = pv.StructuredGrid(X, Y, Z)
mesh.point_data['values'] = data.ravel(order='F') 
pl = pv.Plotter()
pl.add_mesh(mesh.outline(), color="k")
pl.add_mesh(mesh.contour(isosurfaces=np.linspace(1000, 10000, 10)), opacity=0.5, clim=[1000, 10000])
pl.show_bounds(grid='back', location='outer', ticks='inside', zlabel='level', bold=False, font_size=12.0)
pl.show()

Now I’m looking at converting that over into pure VTK and I’ve got something mostly working using a vtkImageData instance and a vtkContourFilter, as well as a vtkCubeAxesActor.

image
image861×706 115 KB

What I’m having some trouble with is:

  1. The different colors for different levels of the contour seen in the Pyvista example
  2. The legend showing what color corresponds to what value
  3. The camera in my VTK solution is centred around that corner of the vtkCubeAxesActor, rather than the centre of the sphere like it is in Pyvista.

I’ve had a dig around the Pyvista source code but it’s got me scratching my head, any advice would be appreciated.

I’ve included my VTK code below:

colors = vtkNamedColors()

axis1Color = colors.GetColor3d("Salmon")
axis2Color = colors.GetColor3d("PaleGreen")
axis3Color = colors.GetColor3d("LightSkyBlue")

x = np.array(np.arange(-steps_range,steps_range,steps))
y = np.array(np.arange(-steps_range,steps_range,steps))
z = np.array(np.arange(-steps_range,steps_range,steps))
data = np.array(calculateData(x, y, z))

image_data = vtkImageData()
image_data.SetOrigin(0, 0, 0)
image_data.SetSpacing(2, 2, 2)
image_data.SetDimensions(100, 100, 100)

image_data.GetPointData().SetScalars(numpy_to_vtk(data.ravel(order='F')))

image_data.Modified()
iso = vtkContourFilter()
iso.SetInputData(image_data)
iso.GenerateValues(10,1000,10000)

mapper = vtkPolyDataMapper()
mapper.SetInputConnection(iso.GetOutputPort())

actor = vtkActor()
actor.SetMapper(mapper)
actor.GetProperty().SetOpacity( 0.2 )
actor.GetProperty().SetColor(1,0.2,0.2)
  
renderer = vtkRenderer()
ren_win = vtkRenderWindow()
ren_win.AddRenderer(renderer)
iren = vtkRenderWindowInteractor()
iren.SetRenderWindow(ren_win)

cubeAxesActor = vtkCubeAxesActor()
cubeAxesActor.SetUseTextActor3D(1)
cubeAxesActor.SetBounds(image_data.GetBounds())
cubeAxesActor.SetCamera(renderer.GetActiveCamera())
cubeAxesActor.GetTitleTextProperty(0).SetColor(axis1Color)
cubeAxesActor.GetTitleTextProperty(0).SetFontSize(48)
cubeAxesActor.GetLabelTextProperty(0).SetColor(axis1Color)

cubeAxesActor.GetTitleTextProperty(1).SetColor(axis2Color)
cubeAxesActor.GetLabelTextProperty(1).SetColor(axis2Color)

cubeAxesActor.GetTitleTextProperty(2).SetColor(axis3Color)
cubeAxesActor.GetLabelTextProperty(2).SetColor(axis3Color)

cubeAxesActor.DrawXGridlinesOn()
cubeAxesActor.DrawYGridlinesOn()
cubeAxesActor.DrawZGridlinesOn()
cubeAxesActor.SetGridLineLocation(cubeAxesActor.VTK_GRID_LINES_FURTHEST)

cubeAxesActor.XAxisMinorTickVisibilityOff()
cubeAxesActor.YAxisMinorTickVisibilityOff()
cubeAxesActor.ZAxisMinorTickVisibilityOff()

cubeAxesActor.SetFlyModeToStaticEdges()
  
renderer.AddActor(actor)
renderer.AddActor(cubeAxesActor)
renderer.SetBackground(colors.GetColor3d('SlateGray'))
renderer.GetActiveCamera().Azimuth(30)
renderer.GetActiveCamera().Elevation(30)

ren_win.SetSize(640, 480)
ren_win.SetWindowName('BlankPoint')
ren_win.Render()
iren.Start()