PECASE: Perceptual Issues in Data Visualization: Conveying 3D Shape and Depth through Texture

PECASE：数据可视化中的感知问题：通过纹理传达 3D 形状和深度

• 批准号: 9875368
• 负责人: Victoria Interrante
• 金额: $ 47.2万
• 依托单位: University of Minnesota-Twin Cities
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-07-01 至 2005-06-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9875368&HistoricalAwards=false
• 关键词: PECASE; Perceptual; Issues; Data; Visualization

Visualization - the visual representation of information through images -is vital in a broad range of scientific and medical applications. A goodvisualization conveys the essential information in a complex datasetclearly and effectively, emphasizing important features and minimizingextraneous detail. The usefulness of a particular visualization techniquecan be quantified in terms of the extent to which it improves taskperformance for a given application. This makes it possible to comparedifferent methods or to pick optimal parameter values for a given method.The process of designing an effective visualization, however, remains forthe most part a black art, terribly ad hoc, largely guided by intuition.The PI's research is a quest for a deeper understanding of the sciencebehind this art, guided by theoretical and experimental insights from humanvisual perception. This can be used to enhance all computer interfacesthat use elements studied here.Specifically, the focus of the research is to investigateperceptually-based methods for more effectively conveying thethree-dimensional shape and relative depth of smoothly curving surfaces inscientific datasets. Key applications for this work include thevisualization of superimposed isovalue surfaces in 3D scalar fields, andthe combined visualization of external and internal structures in medicineand CAD. The investigator will begin by using controlled observerexperiments to thoroughly evaluate texture as a device for enhancing theperception of shape. This will identify and quantify the key aspects oftexture in interface design. The investigator will also pursue theautomatic generation of non-photorealistic, pen-and-ink stylerepresentations of arbitrary polygonally-defined objects for interactive applications. The advantage of such techniques is that some datasets need to be represented in a moresubtly indicative style than traditional graphics renderings. Complementingthe research component of this project is an educational component,directed at both undergraduate and graduate students, aimed at inspiringinterest and fostering creativity through hands-on experiences invisualization.

可视化——通过图像对信息进行视觉表示——在广泛的科学和医学应用中至关重要。良好的可视化可以清晰有效地传达复杂数据集中的基本信息，强调重要特征并最大限度地减少无关细节。特定可视化技术的有用性可以根据其提高给定应用程序的任务性能的程度来量化。 这使得比较不同的方法或为给定方法选择最佳参数值成为可能。然而，设计有效可视化的过程在很大程度上仍然是一种黑魔法，非常临时，很大程度上由直觉引导。PI 的研究是以人类视觉感知的理论和实验见解为指导，寻求对这门艺术背后的科学有更深入的了解。 这可用于增强使用此处研究的元素的所有计算机界面。具体来说，研究的重点是研究基于感知的方法，以更有效地传达科学数据集中平滑弯曲表面的三维形状和相对深度。这项工作的主要应用包括 3D 标量场中叠加等值曲面的可视化，以及医学和 CAD 中外部和内部结构的组合可视化。研究人员将首先使用受控观察者实验来彻底评估纹理作为增强形状感知的工具。这将识别和量化界面设计中纹理的关键方面。研究人员还将致力于自动生成用于交互式应用程序的任意多边形定义对象的非真实感笔墨风格表示。这种技术的优点是，一些数据集需要以比传统图形渲染更微妙的指示性风格来表示。作为该项目研究部分的补充，教育部分针对本科生和研究生，旨在通过可视化的实践经验激发兴趣并培养创造力。