基于GPU加速的虚拟内窥镜手术实时冒烟仿真

Real-time Smoking Simulation Based on GPU Acceleration for Virtual Endoscopic Surgery

采用计算流体力学建模方法提出了一种实时的动态三维冒烟仿真模型;利用定常、温度不变的不可压缩Navi-er-Stokes方程组构造冒烟仿真模型。为了满足虚拟内窥镜手术训练过程中冒烟仿真的实时性和视觉效果上逼真性的要求,利用基于CUDA+GPU软硬件体系结构的高性能计算平台,通过多线程的并发执行实现冒烟仿真模型的大规模并行计算,并使用基于硬件的光线跟踪算法对烟雾进行绘制。实验结果表明所述方法是有效的。

A real-time three-dimensional dynamic smoking simulation based on Computational Fluid Dynamics （CFD） is proposed. Navier-Stokes equations are adopted to construct the smoking simulation model as a constant and incompressible fluid. In order to meet the real-time and realistic visual effects of smoking simulation in training process of virtual endoscopic surgery, CUDA ＋ GPU technologies are employed to realize the large-scale parallel computing of smoking simulation model, and the Ray-tracing algorithm based on hardware is utilized to implement the rendering of smoke. The experimental results show that our method is effective.