近年来,随着科学数据的快速增长,海量时变数据的分析成为了亟待科学家解决的难题.作为科学可视化领域的一类基本方法,体绘制与特征追踪能够很好地辅助科学家理解数据.文中针对科学数据大规模、时变、异构的特点,从大规模数据的实时体绘制,以及时变、异构数据的特征追踪两个方面对现有的研究工作进行了梳理,分别从不同的角度对相关的方法进行了分类和汇总.最后,对科学可视化领域未来的研究方向进行了展望.
In recent years, with the rapid increase of scientific data, there has been increasing concern about the analysis of large-scale time-varying data. As a basic approach in scientific visualization, feature tracking based on volume rendering helps scientists better analyze and understand the data. According to the characteristics of scientific data: large-scale, time-varying and heterogeneous, this paper does a survey of current approaches and look to the future from two aspects: real-time rendering of large-scale data and feature tracking of time-varying, heterogeneous data.