Re: [Algorithms] Efficient Voxel Interpolation / Splatting
Brought to you by:
vexxed72
From: Jon O. <jon...@gm...> - 2010-06-13 19:11:19
|
If its a 2D image to 3d voxel rasterization algorithm, definitely make a oct-tree of the voxel data and then intersect it with the 2D image plane (plus a depth if you want water-tight), and further clip by the extents of the image. Then for each voxel which is considered intersecting the 2D image volume, determine the color by projecting the voxel center onto the image plane then do a sample (bi-linear interp with mip-maps perhaps?) of the image. You could then alpha the voxel based on the amount that the voxel is intersected with the volume. For example 20% inside = 20% alpha. On Sun, Jun 13, 2010 at 1:56 PM, Jon Olick <jon...@gm...> wrote: > Ah, so kind of like a rasterization of the 2d image into the 3d voxel > space? > > > On Sun, Jun 13, 2010 at 1:31 PM, Manuel Massing <m.m...@wa... > > wrote: > >> Hi, >> >> > @Manuel: I think he wants to find the nearest point on the image to the >> > voxel. I'm not sure how a DDA tracing algorithm would help there. >> >> I was thinking of the following: >> Think of each 2d scanline as a 3d ray which passes through the >> voxel grid. Traverse the voxel grid along the ray direction using >> DDA, Insert the intersecting pixel spans (in fact a quadrilateral) into >> the >> appropriate voxel. >> >> Additionally, one could use summed area tables or somesuch to >> pre-filter/match >> the voxel and pixel footprint approximately. >> >> bye, >> >> Manuel >> >> >> ------------------------------------------------------------------------------ >> ThinkGeek and WIRED's GeekDad team up for the Ultimate >> GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the >> lucky parental unit. See the prize list and enter to win: >> http://p.sf.net/sfu/thinkgeek-promo >> _______________________________________________ >> GDAlgorithms-list mailing list >> GDA...@li... >> https://lists.sourceforge.net/lists/listinfo/gdalgorithms-list >> Archives: >> http://sourceforge.net/mailarchive/forum.php?forum_name=gdalgorithms-list >> > > |