Reduced smoothing, slightly darker filopodia.
Processed the whole sequence, texture coherency preserved with old/naive FF generator.

Every filopodium has its own, actually.

RenderOneTimeTexture() and the whole thing modified so that cell body
looks similar to overexpressing/02 dataset.
TODO: Filopodia not visible at all, coherent texture!

This one uses the ActiveMesh::VolID.
But old version works actually better due to spatial-overshoot...

Rendering of texture was spit into three phases.

Added one new param to ActiveMesh::ImportVTK_Volumetric()

Gaussian-smoothing-based FF creation finished -- ActiveMesh::ConstructFF()

Removed coordinate scaler (100x) and filopodia radii scaler. The latter is now a parameter to the system with default value of 1.0.

Added ActiveMesh::ConstructFF() and ActiveMesh::RotateVolPos().
Finished ActiveMesh::FFDots().
Added GetPixel().

RenderMesh() in main.cpp features coherent texture rendering (but without following mesh deformation...)

Added macro: SAVE_INTERMEDIATE_IMAGES

ISOTROPIC version, first one. Resampled for anisotropic closer to the end.

Main() reads files, or pops up OpenGL.
Importing VTKs made more liberal, and silent.

Batch is producing masks and texture (modified RenderMesh()).

Normals are not used when drawing triangle edges.

Finished tweaking of OpenGL vizual mode (depth buffer on, backfaces culling)

Calculation of the 3rd coordinate seems to work well -- at least for a sphere.

GetClosestPointOnQuadricSurface() is the one.

- testing/highlighted vertex changeable with 'v'
- corrected bug when reading STL meshes

- load STL triangle mesh
- display (somewhat rotate and zoom) the mesh
- connect with files from MotionTracking