search

ZackMisso / TransmittanceEstimation

A set of example tests from my SIGGRAPH Asia 2019 publication: Integral formulations of volumetric transmittance
31 stars 4 forks source link

[SOLVED] Regarding the geometry files (photo_backdrop_smooth_2.ply) #1

Closed mmmovania closed 1 year ago

mmmovania commented 1 year ago

Thanks @ZackMisso for sharing your work with us. I was trying to replicate the results in your paper. I was able to successfully generate the fennec_fox scene as in your paper using current openvdb version.

I am now trying to replicate the sphere_cloud scene. In the globe.cpp file there is reference to photo_backdrop_smooth_2.ply file but I cannot locate it anywhere online. If possible, can you please share the geometry or else suggest how i may replicate the same look by using planes and quads similar to how they are used in initializeWalls function in fox.cpp.

Fig. 10 is using geometry that I cannot locate online. I would appreciate if the URL for this model may please be shared.

Thanks once again for an amazing library and awesome research work.

mmmovania commented 1 year ago

Never mind I managed to generate it using my own code.

    Float wf = 1000;
    Float top = 1000;
    Float bot = -23.8;

    // floor
    initializeTrimesh(std::vector<int>{0, 1, 2, 2, 3, 0},
                      std::vector<Float>{0, 0, 5, 0, 5, 5, 0, 5},
                      std::vector<Float>{-wf, -wf, bot, wf, -wf, bot, wf, wf,
                                         bot, -wf, wf, bot});

    // back wall
    initializeTrimesh(std::vector<int>{0, 1, 2, 2, 3, 0},
                      std::vector<Float>{0, 0, 5, 0, 5, 5, 0, 5},
                      std::vector<Float>{-wf, wf, bot, wf, wf, bot, wf, wf, top,
                                         -wf, wf, top});

And here is the output using ratio tracking estimator. The separation of the floor and back wall is visible but i get similar result as your paper Fig. 9. ratio