A NEW SIMPLIFICATION ALGORITHM FOR COLORED OR TEXTURED POLYGONAL MODELS

International audienceMany applications in computer graphics require complex and highly detailed models.However, the level of detail actually necessary may vary considerably. It is often desirable to useapproximations in place of excessively detailed models to control processing time. A newpolygonal mesh simplification algorithm is presented for colored or textured models based onvertex clustering, and a more accurate error-measuring method for vertex clustering is introduced.The algorithm can produce high quality approximations of polygonal models. It makes adaptivesubdivision of the bounding box in the original model using octree structure and performs vertexclustering in an error range specified by users. The color or texture information defined over themesh can be preserved during simplification by constructing a texture map for the simplifiedmesh. To make a continuous transition between level of detail (LoD) models possible, anefficient interpolating method is also proposed. The efficiency of the algorithm is demonstratedin the experimental results