A Three-Dimesional Model Decimation Method

With the rapid development of technology, large and complex three-dimensional model has been more and more common today, It is a great challenge to the storage,transmission, display and rendering these models with the huge amount of data. Although it has lots of simplified algorithm for model, the extreme simplification of complex meshes has received less interest and remains notoriously difficult, small but visually important details are progressively removed from the input scene. Therefore, this paper unite the domestic and foreign research papers on the simplification algorithms for mesh models, propose a new simplification of surface meshes trying to maintaining important features of the model when extremely simplified model.In this paper, aiming at the problem of over-simplification in simplification algorithms now, a improved QEM algorithm is presented. It can extremely simplify the number of vertices and the number of triangular mesh, while small but visually important details can not progressively remove from the input scene. Using the simplified model replaced the original model will impove the speed of redering the model. improved QEM propose the notion of planar proxies, it plays an important role during simplification to protect the structure and to error metrics. We improve the rule of error metrics from the original QEM alrotithm and propose a new rule of error metrics which combine inner quadric with boundary quadric. We propose the notion of maximal clique and witness to preserve and to detect the corner during the decimation. The improved QEM algorithm also propose some structure-preserving rules to prevent the wrong edge collapse operators which will change the structure of the model. Finally, A model simplification system based on the improved QEM algorithm is designed, it can load model, simplify model and display the model imformation.Experimental results prove the effectiveness of improved QEM algorithm,experiments on a variety of inputs illustrate the potential of our approach in terms of improved accuracy and preservation of structure...