Research issues in scientific visualization

As an emerging research discipline, scientific visualization is developing those trappings that demonstrate growth. New algorithms are just beginning to effectively handle the recurring scientific problem of data collected at nonuniform intervals. Volume visualization today is being extended from examining scientific data to reconstructing scattered data and representing geometrical objects without mathematically describing surfaces. Fluid dynamics visualization affects numerous scientific and engineering disciplines. It has taken its place with molecular modeling, imaging remote-sensing data, and medical imaging as a domain-specific visualization research area. Recently, much progress has come from using algorithms with roots in both computer graphics and machine vision. One important research thread has been the topological representation of important features. Volume and hybrid visualization now produce 3D animations of complex flows. However, while impressive 3D visualizations have been generated for scalar parameters associated with fluid dynamics, vector and especially tensor portrayal has proven more difficult. Seminal methods have appeared, but much remains to do. Great strides have also occurred in visualization systems. The area of automated selection of visualizations especially requires more work. Nonetheless, the situation has much improved, with these tools increasingly accessible to scientists and engineers     

Research issues in the foundations of visualization

We focus on the research required to establish a foundation for the evolving needs of visualization systems. Three main topics underpin these needs: models, the need for abstractions to describe the core components of the visualization process and the interfaces between them, including users and their behavior; validation, the problem of determining whether visualizations meet consistency and effectiveness criteria on test data or measures; and systems, the design, realization, and operational problems of systems integrating a range of functionalities to give scientists a working environment for visualization. We outline key aspects of each topic, commenting on the current status of work and isolating areas that require significant research. We conclude by suggesting strategies to initiate this research     

Research issues in data modeling for scientific visualization

This article summarizes some topics of modeling as they impinge on the future development of scientific data visualization. The benefits from visualization techniques in analyzing data are well established, but to build on these pioneering efforts, one must recognize modeling as a distinct structural component in the larger context of visualization and problem-solving systems. Volume modeling is the entry way to this arena of future development, and model-based rendering describes how scientists will view the results. Important side developments such as multiresolution modeling and model-based segmentation will contribute structural capability to these systems. All of these components ultimately depend on the mathematical foundations of scattered data modeling and on model validation and standards to incorporate this modeling methodology into effective tools for scientific inquiry     

Research issues in model-based visualization of complex data sets

At the most abstract level, data visualization maps discrete values computed over an n-dimensional domain onto pixel colors. It is largely a dimension-reducing process justified by its leverage on human perceptual capacities for extracting information from visual stimuli. The difficulty is to implement a mapping that reveals the data characteristics relevant to the application at hand. Effective visualization solutions let the user control the process parameters interactively and enhance the automatically extracted features. We argue for an intelligent, model-based approach to visualization, which extracts the intrinsic data characteristics and constructs multiresolution graphics models suitable for interactive rendering on commercially available hardware adapters. The model-based approach has four parts, which we summarize     

Direct interval volume visualization

We extend direct volume rendering with a unified model for generalized isosurfaces, also called interval volumes, allowing a wider spectrum of visual classification. We generalize the concept of scale-invariant opacity—typical for isosurface rendering—to semi-transparent interval volumes. Scale-invariant rendering is independent of physical space dimensions and therefore directly facilitates the analysis of data characteristics. Our model represents sharp isosurfaces as limits of interval volumes and combines them with features of direct volume rendering. Our objective is accurate rendering, guaranteeing that all isosurfaces and interval volumes are visualized in a crack-free way with correct spatial ordering. We achieve simultaneous direct and interval volume rendering by extending preintegration and explicit peak finding with data-driven splitting of ray integration and hybrid computation in physical and data domains. Our algorithm is suitable for efficient parallel processing for interactive applications as demonstrated by our CUDA implementation.     

Particle-based non-photorealistic volume visualization

Non-photorealistic techniques are usually applied to produce stylistic renderings. In visualization, these techniques are often able to simplify data, producing clearer images than traditional visualization methods. We investigate the use of particle systems for visualizing volume datasets using non-photorealistic techniques. In our VolumeFlies framework, user-selectable rules affect particles to produce a variety of illustrative styles in a unified way. The techniques presented do not require the generation of explicit intermediary surfaces.     

Importance-driven feature enhancement in volume visualization

This paper presents importance-driven feature enhancement as a technique for the automatic generation of cut-away and ghosted views out of volumetric data. The presented focus+context approach removes or suppresses less important parts of a scene to reveal more important underlying information. However, less important parts are fully visible in those regions, where important visual information is not lost, i.e., more relevant features are not occluded. Features within the volumetric data are first classified according to a new dimension, denoted as object importance. This property determines which structures should be readily discernible and which structures are less important. Next, for each feature, various representations (levels of sparseness) from a dense to a sparse depiction are defined. Levels of sparseness define a spectrum of optical properties or rendering styles. The resulting image is generated by ray-casting and combining the intersected features proportional to their importance (importance compositing). The paper includes an extended discussion on several possible schemes for levels of sparseness specification. Furthermore, different approaches to importance compositing are treated.     

Saliency-guided enhancement for volume visualization

Recent research in visual saliency has established a computational measure of perceptual importance. In this paper we present a visual-saliency-based operator to enhance selected regions of a volume. We show how we use such an operator on a user-specified saliency field to compute an emphasis field. We further discuss how the emphasis field can be integrated into the visualization pipeline through its modifications of regional luminance and chrominance. Finally, we validate our work using an eye-tracking-based user study and show that our new saliency enhancement operator is more effective at eliciting viewer attention than the traditional Gaussian enhancement operator.     

Towards animating ray-traced volume visualization

Ray-tracing volumetric data may take several minutes to compute a single image from a fixed viewpoint. We present techniques that generate approximate ray-traced volumetric images in less than one second per image, after a lengthy initialization process is performed. These approximate images are based on methods that interpolate data sampled at locations on a sphere.     

Direct visualization of volume data

A combination of segmentation tools and fast volume renderers that provides an interactive exploration environment for volume visualization is discussed. The tools and renderers include mechanisms that distribute volume data across multiple processors, as well as image compositing techniques and solutions to representation problems in the selection and display of subregions within bounding volumes. A volume visualization technique using the interactive control of images rendered directly from volume data coupled with a user-controlled semantic classification tool is described. The variations of parallel volume rendering being explored on the Pixel-Planes 5 system and the region-of-interest selection methods and the interactive tools used by the system are presented. The flexibility and power of combining volume rendering with region-of-interest selection techniques are demonstrated using examples of medical imaging applications     

大规模结构网格数据的相关性统计建模轻量化方法

高置信度的数据可视分析对于大规模数值模拟至关重要,但是当前高性能计算机的存储瓶颈导致可视分析应用获取原始高分辨率网格数据越来越困难. 基于统计建模的方法能够极大降低高分辨数据存储成本,但是重建数据的不确定性高. 为此,提出了一种大规模结构网格数据的相关性统计建模轻量化方法,用于对并行数值模拟生成的大规模多块体数据进行高效分析与可视化. 该方法的技术核心是通过数据块间的统计相关性,指导邻接数据块的统计建模,从而有效地保留数据统计特征,且不需要对不同并行计算节点中的数据块进行合并与重新分块. 通过耦合数据块的数值分布信息、空间分布信息和相关性信息,该方法可以更精确地重建原始数据,降低可视化的不确定性. 实验测试采用了最大10亿网格规模的5组科学数据,定量分析结果显示,在相同数据压缩比下,该方法相比现有方法可将数据重建精度最大提升近2个数量级.

     

Creative parameter selection for volume visualization

Most of the available algorithms for scalar volume visualization offer predefined techniques such as display of volumetric regions defined by scalar threshold values. The regions can usually be drawn opaque or transparent or appear in combinations. This paper presents an implementation of a volume visualization concept where several modelling and rendering techniques can be applied in any combination, mainly bounded by the creativity of the user. The concept is based on the use of a model for light scattering in a field of varying density emitters, and the use of fixed visual references to improve readability and disambiguate interpretation. An image is computed by means of an interval-based mapping from scalar range to visual parameters. Each interval has a set of associated parameters, such as colour and attenuation. In addition, each interval of the scalar range must be mapped into relative density by means of a transfer function which is selected in a ‘natural way’ depending on application. A methodology is suggested which enable the piecewise transfer functions to be easily determined. A prototype user-interface for the mapping from scalar values to visual parameters is demonstrated. The interface is easy to use for the beginner, while it also encourages creativity and intuition in the process of selecting parameters.     

Uncertainty visualization using HDR volume rendering

In this paper, we explore a novel idea of using high dynamic range (HDR) technology for uncertainty visualization. We focus on scalar volumetric data sets where every data point is associated with scalar uncertainty. We design a transfer function that maps each data point to a color in HDR space. The luminance component of the color is exploited to capture uncertainty. We modify existing tone mapping techniques and suitably integrate them with volume ray casting to obtain a low dynamic range (LDR) image. The resulting image is displayed on a conventional 8-bits-per-channel display device. The usage of HDR mapping reveals fine details in uncertainty distribution and enables the users to interactively study the data in the context of corresponding uncertainty information. We demonstrate the utility of our method and evaluate the results using data sets from ocean modeling.     

Ray casting architectures for volume visualization

Real-time visualization of large volume data sets demands high-performance computation, pushing the storage, processing and data communication requirements to the limits of current technology. General-purpose parallel processors have been used to visualize moderate-size data sets at interactive frame rates; however, the cost and size of these supercomputers inhibits the widespread use for real-time visualization. This paper surveys several special-purpose architectures that seek to render volumes at interactive rates. These specialized visualization accelerators have cost, performance and size advantages over parallel processors. All architectures implement ray casting using parallel and pipelined hardware. We introduce a new metric that normalizes performance to compare these architectures. The architectures included in this survey are VOGUE, VIRIM, Array-Based Ray Casting, EM-Cube and VIZARD II. We also discuss future applications of special-purpose accelerators     

Multimodality visualization of medical volume data

New developments in 3-D volume acquisitions are creating a rapidly increasing demand for integrating multimodality 3-D visualization. In order to accomplish routine clinical multimodality visualization, many issues have to be dealt with, such as techniques for accurate spatial registration, integrated representation, suitable graphical user interfaces, and obtaining adequate rendering speeds. The aim of this experience paper is 2-fold. First, it presents various results from our research on multimodality visualization/registration. Second, this paper explicitly addresses practical problems and findings related to software development and multimodality registration/visualization. We hope that this will give colleagues a better understanding in some of these issues based on our experience, including notably our mistakes.     

Interactive ray tracing for volume visualization

Presents a brute-force ray-tracing system for interactive volume visualization. The system runs on a conventional (distributed) shared-memory multiprocessor machine. For each pixel, we trace a ray through a volume to compute the color for that pixel. Although this method has a high intrinsic computational cost, its simplicity and scalability make it ideal for large data sets on current high-end parallel systems. To gain efficiency, several optimizations are used, including a volume bricking scheme and a shallow data hierarchy. These optimizations are used in three separate visualization algorithms: isosurfacing of rectilinear data, isosurfacing of unstructured data, and maximum-intensity projection on rectilinear data. The system runs interactively (i.e. at several frames per second) on an SGI Reality Monster. The graphics capabilities of the Reality Monster are used only for display of the final color image     

Interactive clipping techniques for texture-based volume visualization and volume shading

We propose clipping methods that are capable of using complex geometries for volume clipping. The clipping tests exploit per-fragment operations on the graphics hardware to achieve high frame rates. In combination with texture-based volume rendering, these techniques enable the user to interactively select and explore regions of the data set. We present depth-based clipping techniques that analyze the depth structure of the boundary representation of the clip geometry to decide which parts of the volume have to be clipped. In another approach, a voxelized clip object is used to identify the clipped regions. Furthermore, the combination of volume clipping and volume shading is considered. An optical model is introduced to merge aspects of surface-based and volume-based illumination in order to achieve a consistent shading of the clipping surface. It is demonstrated how this model can be efficiently incorporated in the aforementioned clipping techniques.