Abstract: To choose the best viewpoint of 3D visualization, the commonly used artificial temptation usually leads to problems of increase of iteration number and poor design efficiency. Therefore, the algorithm based on particle swarm optimization is proposed for viewpoint optimization. The algorithm expresses the viewpoint by using multi-resolution level, and introduces image information entropy for evaluating the quality of the rendered 3D image at different viewpoints. The value of entropy is as the basis of viewpoint optimization and the fitness value of particle swarm optimization. It achieves the intelligent and automatic optimization for viewpoints from the 3D visualization by using the particle swarm optimization algorithm, and thus to realize selection of best viewpoints. The experimental results show that the algorithm has a faster convergence rate, and effectively reduces the number of evaluations. It can improve the quality and efficiency of the rendered 3D image.

Key words: 3D visualization, viewpoint optimization, Particle Swarm Optimization（PSO）, image information entropy

摘要： 通常三维可视化的最佳视点选择是通过人工试探，这样会导致反复迭代尝试的次数增加且效率低下，针对上述问题提出了一种基于粒子群的视点优化方法。该方法把视点利用多分辨率层级来表示，引入图像信息熵评价不同视点下绘制的三维图像的质量，熵值作为视点优化的依据和粒子群的适应度函数值。在三维可视化中利用粒子群算法进行视点的智能、自动的优化，从而实现最佳视点的选择。实验结果表明，该方法具有较快的收敛速度，有效地减少了评估次数，可提高三维可视化的绘制图像的质量和绘制效率。

关键词: 三维可视化, 视点优化, 粒子群算法, 图像信息熵