| Summary: | 碩士 === 國立暨南國際大學 === 資訊工程學系 === 91 === In the research field of computer graphics, the L-system is the mainstream method for modeling plants. There are lots of highly “plants-like” images reported as the research result in many previous works. However, to simulate the interactions between plants and environment, including phototropism, space competition etc., the L-system is still an inconvenient tool. The main purpose of this research is not only to synthesize realistic plant images, but more important is to simulate the interaction between plants and environment. Our basic concept is to combine the traditional L-system and particle system. By applying the particle system to replace the extremely complex parameters, our L-system is more powerful and controllable.
Every alphabet in L-system presents a basic component. The smallest component may be a cell. On the other hand, the largest component may be the root, a leaf, or a stem. Since L-system treats plants as an object composed by many components, it is very easy and intuitive to integrate L-system and particle system together.
To present the multiplicity of plants, our system is basically a non-deterministic <k, l> L-system. For the experiment convenience, we choose some Taiwanese common plants to construct our L-system growth grammar. Because we import the concepts of particle system, every alphabet in grammar is given independent attributes such as its relative position, initial velocity, life cycle, growing rate, color, texture, etc. This approach makes our system be different with the traditional L-system methods.
After the basic plants structure is determined by L-system, every alphabet is treated as a particle generator of particle system. The advantage of importing particle system is that we can obtain adequate control to every plant component. Thus we are able to systematically simulate the interaction between particles and between particles and environment, with the botanic viewpoint such as the foregoing phototropism, space competition, etc. By simulating the interactions with other particles and environment, such as collision detection, shadowing, etc, two plants with the same L-system structure may probably represent quite different appearances in our system.
By integrating the particle system and L-system methods, our plants growth simulation system successfully synthesizes realistic and botanic plants images. We hope our research result will be more contributive to the work of natural conservation by appealing to people for being more comprehended and conscious of the precious nature eco-system through our system.
|
|---|
