面向感知增强的沉浸式流场可视化及交互技术研究

The development of virtual reality technology has driven to a widespread application of immersive flow field visualization technology. However, the inadequacies of traditional one-channel precise interaction mode such as the mouse, keyboard and touch in an immersive virtual environment limit the efficiency and accuracy of flow visualization. This project aims at the complex flow field data and combines flow visualization and human-computer interaction technology to explore a complete visual flow of perception enhanced immersive flow visualization. This project focuses on the visual perception enhanced flow field rendering algorithms, which can effectively display clear flow field patterns and important flow characteristics and improve the quality of visualization rendering. The research on visualization design of flow field based on cognitive efficiency enhancement is investigated in this project. A three-dimensional perception model is established on the multi-dimensional time-varying flow field data to enhance the visual perception of the flow field data. The immediate feedback enhanced immersive interactive technology is also studied in this project. We create a realistic virtual environment by using the head-mounted display device. Combining with the operational characteristics and special requirements in the visual rendering of the immersive flow field, the interaction paradigm is redefined to improve the user's immersive experience and realize interactive characteristic analysis. It provides the users an immersive flow field visualization environment with simple operation, strong applicability, good expansibility and high personal experience, and it can effectively improve the visual quality of the scientific data in the flow field and the efficiency of analysis and comprehend.

虚拟现实技术的火热发展带动沉浸式流场可视化技术的广泛应用，而基于鼠标、键盘以及触控等传统单通道精确交互方式在沉浸式虚拟环境中的局限性，影响流场可视化的分析效率与准确性。本项目面向复杂流场数据，结合流场可视化与人机交互技术，探索构建一个完整的面向感知增强的沉浸式流场可视化流程。研究基于视觉感知增强的流场可视化绘制算法，有效展现清晰的流场模式和重要的流动特性，提高可视化绘制质量；研究基于认知效率增强的流场可视化呈现设计，对多维时变流场数据建立立体感知模型，加强对流场数据的视觉认知；研究基于直觉反馈增强的沉浸式交互技术，借助头戴式显示设备营造逼真的虚拟环境，结合沉浸式流场可视化绘制中的操作特点与关注需求，重新定义交互范例来改善用户的沉浸式体验，实现交互式特征分析。为用户提供了操作简单、适用性强、扩展性好、个人体验度高的沉浸式流场可视化环境，有效提高流场科学数据的可视化质量及分析和理解效率。

本课题以增加用户对流场物理过程感知为目标，结合沉浸式模拟在空间沉浸感、用户参与度、多维感知等方面的分析优势，探索构建一个完整的面向感知增强的大规模沉浸式流场可视化流程。在绘制算法方面提出了一种涡流结构特征增强的三维流线及其衍生绘制技术，引入三维流场矢量特性和拓扑结构特征来提高流线视觉效果和属性表征能力。提出一种基于稀疏噪声纹理生成的改进三维矢量场VolumeLIC绘制技术，有效解决由于三维矢量场本身的空间特性造成的纹理片元之间视线遮挡和覆盖问题。以头戴式显示设备+体感控制器构建沉浸式分析的硬件实现平台，通过空间坐标变换将绘制结果映射到头戴式显示设备构建的三维虚拟空间中。结合人机交互基本理论以及沉浸式分析的技术特点，定义沉浸式交互技术应该具备的基本属性，以基本属性为指导，定义一系列用于支持沉浸式分析的交互设计指南。以所见即所得的非接触手势交互理念为依据，定义符合用户自然行为习惯的操作手势，支持对沉浸式场景下体可视化和流线可视化绘制结果的光学属性、结构参数和观察视角进行直接操作。定义基于手势动作轨迹的客观评估模型—GT(Gesture & Task)模型，结合运动轨迹测量法与绩效测量法两种评估方法，评估手势交互设计的有效性和可用性。引入多视图协同交互模型实现对数据属性特征统计、临界状态分布和时空演化规律的多维度并行呈现。构建了支持多维关联对比的流场可视分析具体实施方案，支持整体-细节循环探索的交互分析流程，借助并列、重叠、显示编码等比较可视化技术呈现不同参数环境下的视觉结果，实现对多维属性数据的全方位、多角度观察、对比与分析。设计并开发了大规模CFD流场可视化分析系统FVIS，系统具有完善而实用的大规模三维定常和非定常流场数据预处理与特征提取、体绘制、几何图形（流线、等值面等）绘制、纹理绘制等功能，并提供多种数据分析手段和良好的交互方法。通过8种展示方式、3种显示模式以及5种视图类型展示流场数据内外流结构特征及其复杂的物理现象。