CAJ | 학술논문

三维数据场可视化是目前计算机图形学方向研究的热点之一，随着人们需求的不断提高，对图形绘制的实时性及交互性提出了更高的要求。本文主要阐述了在CUDA架构下，利用GPU的多核并行运算能力，对实现三维数据场可视化的光线投射算法进行优化和改进。将图形处理的几何阶段在可编程硬件实现，实现图形绘制的加速。实验结果表明，在一定程度上改善了成像效果，绘制速度的加速比达到7左右。
삼유수거장가시화시목전계산궤도형학방향연구적열점지일，수착인문수구적불단제고，대도형회제적실시성급교호성제출료경고적요구。본문주요천술료재CUDA가구하，이용GPU적다핵병행운산능력，대실현삼유수거장가시화적광선투사산법진행우화화개진。장도형처리적궤하계단재가편정경건실현，실현도형회제적가속。실험결과표명，재일정정도상개선료성상효과，회제속도적가속비체도7좌우。
Visualization of 3D data set is one of the research hotspots of computer graphics. With the improve-ment of people’s needs, higher requirements should be made on the real-time rendering and interaction. This paper mainly expounds the optimization and improvement on realizing 3D data set visualization ray casting algo-rithm with CUDA and multicore parallel computing ability of GPU in order to accelerate graphics rendering by implementing graph processing of geometric phase in programmable graphics hardware. The experimental results show that the imaging effect is improved to some extent and that the rendering speed ratio reaches about 7.