Modeling and Visualization of Flower Color Patterns

Flowers are familiar in virtual scenes, however, the design of flower patterns is still mainly done by hand. To produce a number of flower color patterns required by a large scene can be very labor-consuming. In this paper, we present a biologically-motivated algorithm for modeling and visualization of flower color patterns. It is able to produce various flower color patterns with little user interaction. In our system, pigmentation is simulated by a modified reaction-diffusion system, and the simulation function is controlled by a few experiential parameters. Thus it can be adjusted to produce various flower color patterns widely observed in the real world. Furthermore, our algorithm can be easily embedded into other advanced shading models to improve the quality of their rendering results. We present an image-based texture generation method as an example.