A method for moving trees realistically in wind.

The place for artists, modellers, level designers et al to discuss their approaches for creating content for OGRE.
epha
Halfling
Posts: 51
Joined: Tue Dec 11, 2007 11:34 pm

A method for moving trees realistically in wind.

Post by epha »

Hi.
I have been writing an implementation of the method described in Habel, Kusternig and Wimmer's "Physically Guided Animation of Trees" to animate trees in 'wind' in real time. It is not finished but the basics are in place and it works. I'm posting here in case anyone wants to use it as it is, or improve it, and so if there are any problems I don't know about I can find them and fix them. I will continue to work on it but my graphics card limits what I can do at the moment. It takes a 4 step approach to getting trees in Ogre and swaying realistically in the wind.

1. Create a tree in Blender using Gen4 python script ( modified from Sergey Prokhorchuk (Stager)'s Gen3 python script. [ This exports an xml file ]
2. Export tree mesh using Ogre's BlenderExporter.
3. Run the Gen4_TreeData_Convertor on the xml file exported from Gen4 Blender python script.
4. Use a vertex shader to move every vertex based on a vertex texture fetch (you need a card that can do vtf) of the data and motion textures produced by Gen4_TreeData_Convertor.

Requirements:
-Blender
-Ogre + Exporter for Blender
-to build Gen4_TreeData_Convertor:
fftw3
OpenEXR
boost
TinyXml
I need to add support for leaves in the python script and the Gen4_TreeData_Convertor, and clean up the code a bit,
but if anyone wants to use it or improve it download at:http://zeroplusequals.net/downloads/Gen4.zip

This includes: -Gen4 python scripts for Blender
-source code for Gen4_TreeData_Convertor and simple SConstruct for building with scons
-glsl shaders ( obviously needs to be re written in cg or hlsl if that's what you use)
-basic Gen4.material
--------------------------------------------------
The tutorials for Gen3 are hard to find so I put them up here:http://zeroplusequals.net/downloads/Gen3_tuts.zip

epha
Last edited by epha on Sat Sep 12, 2009 6:27 pm, edited 1 time in total.
sunrisefe
Halfling
Posts: 42
Joined: Mon Oct 06, 2008 7:43 am

Re: A method for moving trees realistically in wind.

Post by sunrisefe »

Thanks for your contribution very much.
I download the Gen4.zip but didn't find any executable file, btw, can it be run under microsoft visual studio 2005? Thank you.
User avatar
Juan
Gnoblar
Posts: 23
Joined: Sat Aug 22, 2009 2:55 am
Location: Sevilla, Spain

Re: A method for moving trees realistically in wind.

Post by Juan »

I´m going to try it, thanks :wink:
epha
Halfling
Posts: 51
Joined: Tue Dec 11, 2007 11:34 pm

Re: A method for moving trees realistically in wind.

Post by epha »

I download the Gen4.zip but didn't find any executable file, btw, can it be run under microsoft visual studio 2005?
Look at the post: I said there is source code and a SConsruct file for building the Gen4_TreeData_Convertor. There is a readme that explains it too. I work on mac and linux so I don't know about MVS but I'm sure it can be built with that if everything's in the right place.

I forgot to include TinyXMl in the requirments.

epha
epha
Halfling
Posts: 51
Joined: Tue Dec 11, 2007 11:34 pm

Re: A method for moving trees realistically in wind.

Post by epha »

I should also add: before exporting the tree mesh from Blender do ctrl-N to recalculate normals outside.
intelinc
Gnoblar
Posts: 1
Joined: Mon May 17, 2010 10:38 am

Re: A method for moving trees realistically in wind.

Post by intelinc »

Hi. It seems these files are no longer available at the above URL. Is there somewhere else I might be able to obtain them? I'm quite interested in the work you've done integrating the concepts from the Habel et. al. paper. Thanks.

ryan
soulpower
Kobold
Posts: 32
Joined: Wed May 07, 2008 8:23 am
Location: hangzhou,China

Re: A method for moving trees realistically in wind.

Post by soulpower »

It seems these files are no longer available at the above URL. Is there somewhere else I might be able to obtain them? I'm quite interested in the work you've done integrating the concepts from the Habel et. al. paper. Thanks.



55555 the same reply I want to plaste here
the year of 2009? Is that too long ago?