论文部分内容阅读
提出了一种适合复杂三维建筑物模型高效网络传输的线性几何编码方法。首先对建筑物模型进行体元化,构建原始模型的八叉树索引,然后依据索引结构的层间扩展模式和各模式的经验概率对其几何信息进行线性压缩编码,最后通过构建空间密度场,采用改进的MC算法对线性码流进行解码重构。实验表明,该方法不受原始模型表面拓扑连接关系的约束,能够在保持模型外观的前提下有效减少复杂三维建筑物模型网络传输的数据量和等待时间。
A linear geometric coding method suitable for efficient network transmission of complex 3D building models is proposed. Firstly, the building model is voxelized, and the octree index of the original model is constructed. Then the geometric information is linearly compressed and encoded according to the inter-level extension mode of the index structure and the empirical probability of each mode. Finally, by constructing the spatial density field, The improved MC algorithm is used to decode and reconstruct the linear code stream. Experiments show that this method is not constrained by the topological connection of the original model and can effectively reduce the amount of data and waiting time of the network transmission in the complex three-dimensional building model while maintaining the appearance of the model.