Illustrative Graphics & Visualization Framework for Geological-Inspired Conceptual Model Development

在任何建模活动的早期阶段，用户协作构建一系列概念模型，这有助于制定想法、提出可能的解决方案并做出影响后续建模阶段的重要决策。概念模型通常以手工绘制的草图和渲染图的形式表示，为探索不确定性提供丰富的视觉参考，并帮助描绘更具解释性的、近似的预期模型版本。在此过程中为帮助用户而提出的计算工具的基础是基于草图的界面和建模 (SBIM)、非真实感渲染 (NPR) 和说明性可视化 (IVis) 的计算机图形和可视化领域。这些领域引入了一种新的范式，利用传统插画家的基础和应用绘图和渲染技能，使用户能够更直观地构建和可视化模型，并获得探索其不确定性的新见解。 我建议解决 SBIM 中的基础研究问题，NPR/IV 是在我之前将交互式建模和可视化应用于地球科学的工作中确定的；也受到传统地质素描、渲染和插图实践以及我多年来观察到的例子的启发。我提出一个研究计划，其中包含三个相互关联的主题和问题，如下： (A) “通过素描增强科学观察” A1。用于自然结构概念模型开发的逐步观察草图。同步 SBIM 和 NPR，在应用手势绘图和渲染时传输我们观察到的内容，而无需或最少查看草图。 A2。基于规则的自然结构草图和建模。基于自然结构形态及其界面的算法规则，以确保在草图输入和模型输出期间生成真实的结构。 (B)“天然材料和结构的视觉抽象” B1。天然材料的交互式非真实感渲染。以不同的多尺度几何形状和结构关系呈现的织物图案的交互式程序 NPR。 B2.复杂自然结构的交互式说明性可视化。更高级别的渲染抽象受到八种通用科学插图技术和地球科学插图中使用的具体示例的启发。 (C)“基于相似性的概念草图和渲染。”集成的数据和视觉分析算法集，可对现有草图和渲染进行分类，并根据输入模型的多维属性的相似性生成新的草图和渲染。 研究成果可扩展到更广泛的科学和工程学科以及需要以我们受地质启发的解决方案的风格进行概念草图和渲染的应用。

In the early stages of any modeling activity, users collaboratively build a range of conceptual models, which help formulate ideas, suggest possible solutions, and make important decisions that affect subsequent modeling stages. Conceptual models are commonly represented as hand-made sketches and renderings, providing rich visual references for exploring the uncertainties and helping on the depiction of a more interpretive, approximated version of the intended model. Computational tools proposed to aid the user in this process have their foundations in the computer graphics and visualization fields of sketch-based interfaces and modeling (SBIM), non-photorealistic rendering (NPR), and illustrative visualization (IVis). These areas introduced a new paradigm that leverages the traditional illustrator's fundamental and applied drawing and rendering skills, allowing users to build and visualize models more intuitively and to gain new insights into exploring their uncertainties. I am proposing to address fundamental research problems in SBIM, NPR/IVis identified in my previous work on applying interactive modeling and visualization to geoscience; and also inspired by examples of traditional geological sketching, rendering, and illustration practices and examples I observed over the years. I am proposing a research program with three interrelated themes and problems as follows: (A) “Enhancing Scientific Observations via Sketching” A1. Stepwise observational sketching for conceptual model development of natural structures. Synchronized SBIM and NPR, transferring what we observe without or minimally looking at the sketch while applying gestural drawing and rendering. A2. Rule-based sketching and modeling of natural structures. Algorithmic rules based on the morphology of natural structures and their interfaces, to ensure realistic' structures are generated during sketch-input and in the model output. (B) “Visual Abstractions of Natural Materials & Structures” B1. Interactive non-photorealistic rendering of natural materials. Interactive procedural NPR of fabric patterns rendered in varying multi-scale geometries and structural relationships. B2. Interactive illustrative visualization of complex natural structures. Higher-level rendering abstraction inspired by eight general scientific illustration techniques and specific examples used in geo-scientific illustrations. (C) “Similarity-based Conceptual Sketching & Rendering.” Integrated set of data and visual analytics algorithms to classify existing sketches and renderings and generate new ones based on similarity of multi-dimensional attributes from the input models. The research outcomes are extendable to a broader number of scientific and engineering disciplines and applications requiring conceptual sketching and rendering in the style of our geology-inspired solutions.