基于GPU的交互式体数据切割  

GPU-Based Interactive Volume Cutting

作　　者：葛军[1] 姜小明[1] 舒华忠[1] 

机构地区：[1]东南大学影像与科学技术实验室,南京210096

出　　处：《计算机辅助设计与图形学学报》2010年第3期428-433,共6页Journal of Computer-Aided Design & Computer Graphics

摘　　要：为实现实时高效的体数据切割操作,提出一种将人机交互和基于GPU的体绘制相结合的切割方法.先通过人机交互生成多个切割体的几何形状并三角化,将其作为基于GPU的光线投射算法的代理面,然后利用深度剥离算法获得光线在所有代理面上的出入射点坐标,并用多个通道完成体绘制.该方法在单个3D纹理上进行,可以实现多个任意形状的切割体切割,在节省GPU内存空间的同时,提高了切割后的绘制效率.In order to realize real-time and efficient volume cutting, this paper presents a cutting method based on human-computer interaction and GPU volume rendering. By the method, first, the 3D cutting volume surfaces are generated by human interactions and trianglization. These 3D surfaces are used as proxy geometry for GPU-based ray-casting algorithm. Then, the ray＇s enter and leaving points on the proxies are determined through the depth peeling algorithm. Finally, the n-pass volume rendering is achieved. The proposed method is applied to single 3D texture, it achieves arbitrary- shaped multiple cutting. The rendering efficiency after cutting can be improved while the memory on GPU is saved.

关 键 词：体绘制 图形处理器 3D纹理 人机交互 体切割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]