Real-time Simulation For Fish Locomotion Based On Biological Modal And GPU Acceleration

3D computer animation is a very important research field in virtual reality. It simulates the real 3D spaces by computer, constructs 3D model, sets color and texture to its surface, designs locomotion and deformation of the model, designs the color, power, position and motion of lights, designs shooting angle of virtual camera, and at last creates continuous images which is playable.3D animation could create special effects that don't exist in real world. Because it has so many specialties, it is widely applied in many fields.My main work is to simulate fish locomotion and behavior on ordinary PC. Due to the problems existing in current system of real-time simulation on fish locomotion, such as heavy calculation, high complexity and difficulty in real-time simulation on ordinary PC, I propose a new real-time simulation method based on biological model and hardware acceleration. This method, based on the analysis of fish biological modeling and its locomotion, establishes the modularized model of fish, and in the meantime breaks down fish lcomotions into several parts and designs a uniform function for fish locomotions. Finally, through the parallel acceleration processing by GPU, we accelerate the algorithms and realize the real-time simulation of fish's locomotion system. In the mean time, fish are autonomous agents, so a behavior system for fish which could let it to select an appropriate action autonomously is necessary, for example, when to turn around, when to take a rest, how to avoid the obstacles and so on, and I realize those behaviors. Some scene appurtenances are added for a more realistic effect, such as float grass, rockery, air bubbles and so on. Through the methods I mentioned above, I realize the real-time simulation of fish locomotion and behavior on ordinary PC, and achieve a good effect.In the preface of my paper I introduce the technology of computer animation and related works in fish simulation. In the first section, development platform and technologies which I used are mentioned. The following section introduces system overview and scene appurtenances, then the locomotion and behavior system of virtual fish are listed in detail in the next section. In the fifth section I explain the methods I used in collision detection and avoidance. The last part of my paper is conclusion and expectation.