Splatting Without The Blur

Abstract: Splatting is a volume rendering algorithm that combines efficient volume projection with a sparse data representation: Only voxels that have values inside the iso-range need to be considered, and these voxels can be projected via efficient rasterization schemes. In splatting, each projected voxel is represented as a radially symmetric interpolation kernel, equivalent to a fuzzy ball. Projecting such a basis function leaves a fuzzy impression, called a footprint or splat, on the screen. Splatting traditionally classifies and shades the voxels prior to projection, and thus each voxel footprint is weighted by the assigned voxel color and opacity. Projecting these fuzzy color balls provides a uniform screen image for homogeneous object regions, but leads to a blurry appearance of object edges. The latter is clearly undesirable, especially when the view is zoomed on the object. In this work, we manipulate the rendering pipeline of splatting by performing the classification and shading process after the voxels have been projected onto the screen. In this way, volume contributions outside the iso-range never affect the image. Since shading requires gradients, we not only splat the density volume, using regular splats, but we also project the gradient volume, using gradient splats. However, alternative to gradient splats, we can also compute the gradients on the projection plane, using central differencing. This latter scheme cuts the number of footprint rasterization by a factor of four, since only the voxel densities have to be projected. Our new method renders objects with crisp edges and well-preserved surface detail. Added overhead is the calculation of the screen gradients and the per-pixel shading. Both of these operations, however, may be performed using fast techniques employing lookup tables.

We will refer to the traditional splatting method as pre-shaded splatting, since here pre-shaded voxel are projected. The new method is referred to as post-shaded splatting, since shading and classification occurs after the voxels have been projected onto the image or sheet-buffer plane.

Want to know more about it? Here is the paper that will be presented at the 1999 Visualization conference in November in San Francisco.
 

IMAGES:

Post-shaded splatting has a tremendous effect on image quality, as these images readily demonstrate.
 
 

  • Here are renderings of the full UNC head (all images 512 x 512):

    •  
    traditional pre-shaded splatting
    new post-shaded splatting
    pre-shading, but post-classification
    We wanted to see if elements of the pre-shaded pipeline can be combined with elements of the post-shaded pipeline. For the image above on the right we projected pre-shaded voxels, but we also projected their original densities. Then we classified the pixels on the image plane, and set all pixels back to zero where the densities were below the iso-threshold. Doing so causes severe artifacts. See the paper for an explanation.
     
     
  •  The visual differences are especially striking in zoomed views. Here we look the UNC lady in the eye (magification ~5):

    •  
    traditional pre-shaded splatting
    new post-shaded splatting

     
  •  A view onto the brain:

    •  
    traditional pre-shaded splatting
    new post-shaded splatting

     
  •  A zoomed view onto the brain, with a large magnification of ~12:

    •  
    traditional pre-shaded splatting
    new post-shaded splatting
    post-shaded splatting with gradient splats
    So far, we have used central-differences for all post-shaded images. However, for the image above on the right, we used gradient splats. The reward is even crisper detail.
     
     
  •  A view at a ganglion nerve dataset:

    •  
    traditional pre-shaded splatting
    new post-shaded splatting

     

    To Research Homepage