Sunday, January 3, 2010

artSea - the project

It's time to say something more about the project itself. Its goal was stated as follows:
Artificial sea - write a simulation from point of view of a diver moving in a coral reef. Include movement models for inividual fish as well as fish schools (and the movement of other sea animals.)
 We're trying to accomplish this using technologies and algorithms that are well known in game development industry. We currently use:
As for algorithms, we primarily focus on flocking algorithms, as described in Game Programming Gems, volumes I - III.

Currently, we have a working OGRE skeleton application with modified TwOgreGui plugged in and first sketches of flocking algorithm. Finally we want to have at least four different species of marine creatures along with some plant life and coral reef terrain. It's a long road, filled with several obstacles. The main problems we currently face are:
  • High poly models - our freeware fish models converted from 3DS are too hi-poly for real-time applications; we need to find a way to remove some polygons.
  • Performance of simple flocking - we are not sure whether our algorithm can handle a thousant fish at the scene while maintaining acceptable frame rate. It needs to be tested.
  • Nice looking underwater environment - there are some optical effect that might be tricky to implement, as this topic is not well-covered in gamedev literature.
We'll be describing our struggles with those (and other) problems here as the development progresses.

-artSea team-

No comments:

Post a Comment