Monday, October 3, 2011

Reaction: A Survey of Algorithms for Volume Visualization

First, the quote at the beginning “… in 10 years, all rendering will be volume rendering”, is interesting in its perspective as visualization currently (20 years out) does not always use volume visualizations. But beyond that, this paper provides a introduction into the various forms of algorithms that can be used for volume visualizations. This runs the gamut from cube techniques for surface fitting, to projection methods and image-order methods.

I find the argument of photorealism to be very interesting. On the one hand, you have an argument stating that visualization should be limited to what is most commonly interpreted. This is a sound position as it would be easier to interpret a particular data set if it is represented in the most normal format. At the same time, however, information visualization is about abstracting data to provide the most easily interpretable and largest interpretable set of data possible. This lends to the idea that the data being represented should have as much detail in its layers, and abstraction to those layers, for these goals. This could easily mean an abstract visualization technique for volume rendering that includes layered information could be more efficient.