Generalized Parameterized Modeling And Realistic Rendering Of The Trees

The modeling and rendering of tree has been a challenge in computer graphics because of its richness of species, complexity in structure and diversity in shape. The efficiency of tree topology structure expression, quickly and realistic tree rendering has importance significance. According to the output of GreenLab, this paper defines a method for tree structure description based on binary tree, and on this basis we put forward and realize an effective structure simplification method, which not only keeps the visual and biomass fidelity, but also greatly reduces data quantity and improves the rendering efficiency. This paper proposes a generalized parameterized method for tree modeling and builds generalized parameterized modeling and rendering system of trees based on the method. In this system, we besides can model trees according to user input parameters, but also render them realistically and simulate the growth according to growth equations. By employing GPU programming, we optimize the whole system of parameterized modeling and rendering processes, and effectively improve the efficiency of the algorithm.This paper established GreenLab models of poplar 107 and Jing 2(P.×euramericana cl. J2) according to the actual measurement data. We test the whole system by taking these two models as an example, and results show that the generalized parameterized modeling approach helps generate geometric model that can reflect trees growth rule and facilitate the tree growth simulation. The system can not only generate tree morphological structure, which is realistic and botanical correspondent, but also can simulate tree growth smoothly.