一种面向不等尺寸分块海量数据集的并行体绘制算法  

A PARALLEL VOLUME RENDERING METHOD FOR OUT-OF-CORE DATASETS PARTITIONED BY NON-UNIFORM SIZE SUB-BLOCKS

作　　者：朱小烨 薛健 吕科 Zhu Xiaoye;Xue Jian;LüKe(School of Engineering Science,University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院大学工程科学学院,北京100049

出　　处：《计算机应用与软件》2020年第5期44-49,56,共7页Computer Applications and Software

基　　金：国家自然科学基金项目(61671426,61731022,61871258);国家重点研发计划项目(2017YFB1002203);北京市自然科学基金项目(4182071);中国科学院科研装备研制项目(YZ201670)。

摘　　要：传统的体绘制方法需要将原始体数据全部装入内存,无法应用于Out-of-core(海量)数据集。针对这个问题,提出一种有效的针对海量数据集的分块并行体绘制方法,以提高渲染海量数据的速度。在充分区分背景体素的非均匀尺寸分块的体数据集上,通过分块间的遮挡关系构建全局遮挡关系图,对分块进行分组并行渲染。实验表明,该方法在海量数据集上达到了比现有方法更快的渲染速度。The traditional volume rendering method needs to load all the original volume data into memory,which cannot be applied to the Out-of-core datasets.Therefore,we propose an effective sub-blocks parallel volume rendering method for Out-of-core datasets to accelerate the rendering process.It constructed a global occlusion graph by occlusion relationship between sub-blocks on the volume dataset that fully distinguished the non-uniform size of background voxels and rendered the sub-blocks in groups and in parallel.Experimental results show that our method achieves faster rendering speed than the existing methods.

关 键 词：并行体绘制 海量数据集 不等尺寸分块 遮挡关系图 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]