Research On Realistic Grass Rendering Method Based On GPU

In recent years, vegetation rendering is becoming a hot topic in computer graphics. Grass, as an essential element of natural scenes, plays an important role in various fields of applications, such as computer games, landscaping, computer rendering and so on.In previous research, scholars have drawn more attention to grass representation and grass animation, but few studies have focused on the growing area of grass. In the field of computer graphics, the growing area of grass is generally selected by hand s, which definitely arises the wasting of time and effort, and sometimes may appear phenomenon that is inconsistent with the practical situations.In contrast, dividing grass growing region is very popular in biology. However, the research on this field more concentrates on specific vegetation and areas, thus these researches could not be widely used and it is difficult to apply in computer science work directly.To solve above problems, we consider combine biology and computer graphics together. By analyzing the factors that can affect grass growing, the writer of thesis depending the biology knowledge to organize and integrate these factors, to convert them can be used in computer program. Based on genetic algorithm, we can automatically select the grass growing region to achieve grass representation and animation. The detail works list as below:Firstly, by studying the traditional method of grass representation, we can combine the three traditional method of representation together in geometry shader depending on LO D technology and achieve a smooth transition between LO Ds, so, we can simulate the grass field swaying in wind.Secondly, realism of grass depends greatly on whether it is animated or not. After completing the grass representation, we establish the model of wind filed and implement grass animation, which further improves the feeling of realism.Thirdly, based on genetic algorithm theory, we analyze the ecological conditions, weather conditions that can affect grass growing area, and combine the terrain, lighting, water and other data to calculate the best grass growth area in current situation. By doing so, we can complete grass representation and animation in these areas.Experimental results show that the methods we proposed have good results, we provides a new idea for realistic grass rendering, which has certain research and practical value. Meanwhile, the method about grass representation and animation is available to achieve the large scale dynamic grass field.