Based On Non-photorealistic Scene Rendering Technology

Graph is one of the most major media passing on the information, and the development and application of computer graphics are the signs of the development level of computer software and hardware, which is the indispensable importance and foundation stone of information technology. The drawing skills of computer graph are divided into two major types: photorealistic rendering and non-photorealistic rendering. The photorealistic rendering is always the main trend in the development of the computer graphics in over 30 years, but it still has many disadvantages from the artistic and practical perspectives.The non-photorealistic rendering, which makes up for the drawbacks on art and the practical use of traditional photorealistic rendering, is an innovative and active branch of the computer drawing.Firstly, the defects on art and the practical use of the real drawing are analyzed, on the base of it, the thesis does research on relative skills of the non-photorealistic rendering, analyzing and comparing the similarity and dissimilarity between the non-photorealistic rendering and reality.Next, based on the research of non-photorealistic rendering, the thesis comes up with the new way of non-photorealistic rendering with random stroke. Through the place of stroke which depend on the connection between intensity and morphology boundary of the source image's brightness image, the stroke distribution can automatically adapt to the changes with the intensity of source image's details. Computer-assisted instruction system on the stroke is realized and benefits the teaching of computer graphics.Finally, this thesis does the research and analysis on the modern three-dimensional model image edge detection algorithm and image edge drawing algorithm, and sums up the advantages, disadvantages and suitable situation of these algorithms. Based on the analysis above, the thesis puts forward to comprehensive image edge detection algorithm and realizing the fast drawing of three-dimensional model situation.