应用于虚拟内窥镜系统的拓扑细化算法的优化

Optimized Topological Thinning Algorithm for Virtual Endoscopy

虚拟内窥镜技术是近几年迅速发展起来的一种独特的技术。它的关键在于三维物体中心路径的提取。拓扑细化算法可以有效的提取中心路径,但是耗时很长,严重的影响了该算法的实用性。针对这一缺点,作者提出了几种改进,主要通过减少扫描图像的次数和减少体素数目等方法在不同程度上加速算法,提高了算法的效率。实验证明,改进后的算法可以快速稳定的处理体数据,很大程度上起到了优化的作用。

Virtual Endoscopy （VE） is a special technique which has being developed quickly these years, The core of VE is extracting the centerline of the 3D object, Topological thinning is an efficient method to do this job, but the procedure is quite time consuming. The topological thinning algorithm was accelerated using two strategies： First, the times of scan were reduced. Second, the number ofvoxel was reduced, The result of the experiment proves that the optimized algorithm can deal with the data quickly and steadily.