Sign in to follow this  

Procedural Modelling Research - Help!

This topic is 4035 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Recommended Posts

Hi. I'm currently in the beginning stages of an academic project on terrain/procedural modelling. I'm going to be investigating the generation of cavernous terrain. My problem is that I'm having trouble finding resources, and in particular, previous research/work on the topic: I'm not focusing on the geological precision of features in the generated terrains, it's more about how to handle the plotting of geometry in a way that doesn't depend on the usual 1 and 2-dimensional heightmaps, which always displace terrain only vertically. I've had a reasonable amount of experience in terrain programming, and that's why I've chosen a topic like this. I'm having trouble finding the appropriate information on this area of procedural modelling. I know I haven't gone into great detail about what it is I'm trying to do here, but a good example of the techniques I'll need to use can be seen in UnrealEd, when wrapping edges of terrain to create holes and tunnels through hills etc. (above ground in this case, but still the same idea). Before anyone says so, I know that it's perhaps not a vital path of research in terms of where procedural modelling is going, but if anyone can give me any suggestions - papers, sites, journals, or anything you can think of - I'd be grateful. Thanks.

Share this post


Link to post
Share on other sites
Yeah, I've considered them. I should've made it more clear that I'm more concerned about previous work on this subject, so that I can see how this kind of terrain has been built in the recent past. The techniques come afterwards, but if I can't provide a substantial background overview of the subject, I'm not going to get very far.

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
sorry but as near as I know, there isn't any work on *cavernous* procedural terrain.
The majority of work is fractal based as applied to 2d heightmaps... ie no caves

consider that even in non-procedural terrain, caverns are rare and generally limited to first person shooter games where the 'terrain' is typically some stylized portal engine, not an actual terrain in the large scale flight-sim sense

the best bet I can think of would be Voxel based terrains; someone *might* have used 3d fractal fields to distribute voxels and create caverns...

Share this post


Link to post
Share on other sites
Yeah.

I'm aware of the reasons for this type of terrain not being in the mainstream of procedural modelling; anyone can understand that. But it *is* an experiment, and it would be nice to see procedural, semi-random terrain generation stepping up a gear and trying to break out of the height-map syndrome. I'm glad you brought up voxel distribution, though, I forgot about that.

Share this post


Link to post
Share on other sites
Guest Anonymous Poster
Interesting.
The only scientific link I could remember is this one: http://www.noisemachine.com/talk1/8.html

A 'classic' if you will; on first glance this would shift the problem to generating the density distribution.

Concerning this I remember reading about rogue-like games trying to 'grow' cave levels
in 2D; generally something like this: http://www.pixelenvy.ca/wa/ca_cave.html

I tried (not very hard) searching for papers using some of those terms but not much came up.

You probably know this one, but I think it appropriate to link to http://www.vterrain.org/Elevation/Artificial/index.html in this context.

Share this post


Link to post
Share on other sites
I think you need to move to some solid modeling method. BSP trees is the simplest, but there are others (all the way up to Voxels, which are typically too memory intensive to be used at real scale). You might also want to check out the Pixelux DMM technology (I think they partnered with Lucasarts earlier this year): http://www.pixeluxentertainment.com/

Assuming you go with BSP/CSG construction: Once you have the rough shape of the caverns in BSP form, you can start refining the surfaces using detail geometry and tesselation -- I don't think it'd be smart to model every protrusion in the BSP model. This is how games like Unreal work: They start with a large, assumed solid block, from which you "carve" your level. You can also "carve" off the top to generate terrain surface, if you need to. There should be tons of references on this.

Share this post


Link to post
Share on other sites

This topic is 4035 days old which is more than the 365 day threshold we allow for new replies. Please post a new topic.

If you intended to correct an error in the post then please contact us.

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this