基于GPU的三维医学图像混合可视化系统

研究并实现了一个基于GPU的医学图像混合可视化系统，该系统采用三维纹理映射的方法实现直接体绘制，利用GPU的可编程特性完成体绘制方法中的插值后分类算法和传输函数的传递及实时修改，采用OpenGL技术实现表面的绘制，并基于场景图结构实现时表面数据的管理。面绘制和体绘制部分都采用OpenGL实现，运用OpenGL的融合机制，系统实现了面绘制和体绘制的混合显示。本系统大大提高了体绘制的速度，有效地保留了面绘制和体绘制的优势，在保证绘制速度的基础上丰富了图像信息。

3D Medical Image Hybrid Visualization System Based on Graphics Processing Unit

A hybrid visualization system for the medical image is studied. Three-dimensional texture method is used in volume rendering to accelerate the rendering speed and the programmable characteristic of GPU is used to realize the post-classification algorithm and the interactive modification of transfer functions. The surface rendering is realized using OpenGL, while the scene graph structure is used for the tissue management. As OpenGL is used in both surface rendering and volume rendering, it is possible to realize hybrid visualization by using the fusion mechanism of OpenGL. The system has the ability of speeding up volume rendering. The system efficiently uses two rendering techniques and enriches the final image information.