3D visualization of data varying in time - Ideas needed plz!

marachli

Hello everybody..

I'm working on a project at the university about 3D visualization of data varying in time..

I have 40 files (f0.dat ...... f39.dat) each one contains an array (1x1000000) and it represents the values of density for 1000000 voxels at some instant of time.

Example: f5.dat contains the values of density for 1000000 voxels at time t=4 (we start at t=0 till t=39).

I have the 3D coordinates of these voxels in the array coord (3x1000000) and those 1000000 voxels represent a part of a large 3D cube (300x300,300).

I have an idea of detecting the voxel regions that change together in from t to t+1, then to build 3 classes:

classH : for regions that change in high differences from t to t+1
classM: for regions that change in medium differences from t to t+1
classL: for regions that change in low differences from t to t+1

Then I have to separate each class into isolated regions because we need to respect the neighbourhood,a new region from a class will contains the connected voxels that change from t to t+1 in (high/medium/low) differences. The result is a voxel will be a member of one region such as : high1 or medium4 or low3 for example. Then by using libqglviewer I'll visualize by dealing with regions instead of dealing with each voxel.

So please if you have any other ideas that could be better than this one ,I'll be glad to know it.

thanks.
Happy programming!

[fixed formatting -marius]

infoctopus

consider using gnuplot or similar software that can draw 3D plots.
Then a series of plots could be combined to an animated GIF.

andre

Get yourself a copy of "this":http://www.edwardtufte.com/tufte/books_vdqi book, and get some inspiration