RE: [Algorithms] VIPM With T&L - what about roam

[Brian M. <bma...@ra...>]

I've come across 2 papers recently that are worth thinking about with
regards VIPM.

Efficiently Computing and Updating Triangle Strips for Real-Time Rendering
	Jihad El-Sana, Francine Evans, Amitabh Varshney, Steven Skiena, Elvir
Azanli

(Get it from this link, under publications.)
http://www.cs.bgu.ac.il/~el-sana/

	This is more work like strip, including 'Issues in Integrating Triangle
Strips with Multiresolution Hierarchies' which introduces Skip-Strips for
maintaining triangle strips as you do edge collapses. Not specific to VIPM
as it deals with VDPM and Merge Trees, but well worth a look.

Coarse View-Dependant Levels of Detail for Hierarchical and Deformable
Models
	Dieter Schmalsteig, Anton Fuhrmann
(Get it from this link, under publications some way down...)
http://www.cg.tuwien.ac.at/staff/DieterSchmalstieg.html

	This includes a neat way of producing chunked VIPM like Charles Bloom
demonstrates on his web site (and the discussion here.) The interesting bit
is that the authors assign vertices to VIPM clumps and not triangles which
allows for some simplification on the boundary. I won't go into details here
as the authors do a much better job of explaining it than I can.

	Hopefully everyone interested in the VIPM discussions will find these of
interest.

-Brian.

[snip]

> There were some thinking-out-loud discussions between myself and Charles
> Bloom a few weeks ago on a similar subject (doing VIPM with strips rather
> than lists). If the archives don't work, and you don't have the old
messages
> to search through, I could probably dig them up and repost them.
>
> Tom Forsyth - Muckyfoot bloke.
> Whizzing and pasting and pooting through the day.

[snip]