View Dependent Sequential Point Trees

Sequential point trees provide the state-of-the-art technique for rendering point models, by re-arranging hierarchical points sequentially according to geometric errors running on GPU for fast rendering. This paper presents a view dependent method to augment sequential point trees by embedding the hierarchical tree structures in the sequential list of hierarchical points. By the method,two kinds of indices are constructed to facilitate the points rendering in an order mostly from near to far and from coarse to fine. As a result, invisible points can be culled view-dependently in high efficiency for hardware acceleration, and at the same time, the advantage of sequential point trees could be still fully taken. Therefore, the new method can run much faster than the conventional sequential point trees, and the acceleration can be highly promoted particularly when the objects possess complex occlusion relationship and viewed closely because invisible points would be in a high percentage of the points at finer levels.     

结合可见性处理的顺序点模型树

将点模型的树节点按照各自的误差进行顺序排列,就形成了顺序点模型树,这样,就能方便地利用GPU(图形处理器)来高速绘制点模型,但这种方法缺乏对可见性的高效处理,不便于处理大型复杂的模型,因为这些模型中各个部分之间的遮挡关系复杂,且每次成像所需的树节点的误差跨度大,导致了大量不可见的树节点也要绘制,为此,文章提出一种结合可见性处理的顺序点模型树,即在根据误差顺序排列树节点时,融入一些可见性信息,这样,绘制时,不仅能保留顺序点模型树高效利用GPU的优点,还能避免处理大量的不可见但满足误差要求的树节点,实验表明,相比于顺序点模型树,新方法的绘制速度更快,加速率至少在20％,特别是在处理大规模复杂模型时的加速率更高,能达67％以上。     

3D Video Recorder: a System for Recording and Playing Free‐Viewpoint Video†

We present the 3D Video Recorder, a system capable of recording, processing, and playing three‐dimensional video from multiple points of view. We first record 2D video streams from several synchronized digital video cameras and store pre‐processed images to disk. An off‐line processing stage converts these images into a time‐varying 3D hierarchical point‐based data structure and stores this 3D video to disk. We show how we can trade‐off 3D video quality with processing performance and devise efficient compression and coding schemes for our novel 3D video representation. A typical sequence is encoded at less than 7 Mbps at a frame rate of 8.5 frames per second. The 3D video player decodes and renders 3D videos from hard‐disk in real‐time, providing interaction features known from common video cassette recorders, like variable‐speed forward and reverse, and slow motion. 3D video playback can be enhanced with novel 3D video effects such as freeze‐and‐rotate and arbitrary scaling. The player builds upon point‐based rendering techniques and is thus capable of rendering high‐quality images in real‐time. Finally, we demonstrate the 3D Video Recorder on multiple real‐life video sequences. ACM CSS: I.3.2 Computer Graphics—Graphics Systems, I.3.5 Computer Graphics—Computational Geometry and Object Modelling, I.3.7 Computer Graphics—Three‐Dimensional Graphics and Realism     

基于Fork/Join框架的等值面快速生成并行算法

针对传统串行等值面提取算法在处理离散点数量多、网格点密度大的数据时生成效率差的问题,提出一种新的基于Fork/Join框架下的等值面快速生成并行算法。通过对传统串行算法中的关键步骤进行并行计算可行性分析,提出可以实施并行计算的四个单独步骤:离散点数据网格化处理、等值点计算、等值线追踪与光滑、等值面标记识别。通过将并行计算作用于等值面生成的这四个步骤中,减少了等值面计算的执行时间,加快了等值面的生成速度。实验结果表明,在数据计算量较大时,与传统串行算法相比,并行算法能在2秒内快速生成等值面,最大加速比高于5.0,提高了等值面的生成效率并取得了良好的绘制效果,满足了高实时性的业务需求。     

A comparison of task pools for dynamic load balancing of irregular algorithms

Since a static work distribution does not allow for satisfactory speed‐ups of parallel irregular algorithms, there is a need for a dynamic distribution of work and data that can be adapted to the runtime behavior of the algorithm. Task pools are data structures which can distribute tasks dynamically to different processors where each task specifies computations to be performed and provides the data for these computations. This paper discusses the characteristics of task‐based algorithms and describes the implementation of selected types of task pools for shared‐memory multiprocessors. Several task pools have been implemented in C with POSIX threads and in Java. The task pools differ in the data structures to store the tasks, the mechanism to achieve load balance, and the memory manager used to store the tasks. Runtime experiments have been performed on three different shared‐memory systems using a synthetic algorithm, the hierarchical radiosity method, and a volume rendering algorithm. Copyright © 2004 John Wiley & Sons, Ltd.     

On‐the‐fly generation and rendering of infinite cities on the GPU

In this paper, we present a new approach for shape‐grammar‐based generation and rendering of huge cities in real‐time on the graphics processing unit (GPU). Traditional approaches rely on evaluating a shape grammar and storing the geometry produced as a preprocessing step. During rendering, the pregenerated data is then streamed to the GPU. By interweaving generation and rendering, we overcome the problems and limitations of streaming pregenerated data. Using our methods of visibility pruning and adaptive level of detail, we are able to dynamically generate only the geometry needed to render the current view in real‐time directly on the GPU. We also present a robust and efficient way to dynamically update a scene's derivation tree and geometry, enabling us to exploit frame‐to‐frame coherence. Our combined generation and rendering is significantly faster than all previous work. For detailed scenes, we are capable of generating geometry more rapidly than even just copying pregenerated data from main memory, enabling us to render cities with thousands of buildings at up to 100 frames per second, even with the camera moving at supersonic speed.     

大规模点模型的实时高质量绘制

提出了一个针对大规模点模型的实时高质量绘制算法．该算法采用距离相关的自适应绘制策略,在不损失绘制质量的前提下简化了计算．在预处理阶段,对点模型进行剖分,为每一分片分别建立层次结构,并序列化为线性二叉树保存．绘制时,首先根据每片的包围盒和法向等信息进行快速视域裁剪与背面剔除．对于可见的分片,依据它与视点的距离和视线方向,选择恰当的细节层次,直接取出线性二叉树中对应的点几何数据,自适应地选择最合适的绘制模式对其进行绘制．为了减少存储空间的消耗,该文提出了一种面向保留模式图形硬件加速的点模型压缩和解压缩算法,压缩比例达到8：1．作者在可编程图形硬件中实现了该文算法,在普通微机上实现了百万数量级点模型的实时高质量绘制。     

Point-based rendering techniques

The increasing popularity of points as rendering primitives has led to a variety of different rendering algorithms, and the different implementations compare like apples to oranges. In this paper, we revisit and compare a number of recently developed point-based rendering implementations within a common testbed. Also we briefly summarize a few proposed hierarchical multiresolution point data structures and discuss their advantages. Based on a common view-dependent level-of-detail (LOD) rendering framework, we then examine different hardware accelerated point rendering algorithms. Experimental results are given with respect to performance timing and rendering quality for the different approaches. Additionally, we also compare the point-based rendering techniques to a basic triangle mesh approach.     

Compact,Fast and Robust Grids for Ray Tracing

The focus of research in acceleration structures for ray tracing recently shifted from render time to time to image, the sum of build time and render time, and also the memory footprint of acceleration structures now receives more attention. In this paper we revisit the grid acceleration structure in this setting. We present two efficient methods for representing and building a grid. The compact grid method consists of a static data structure for representing a grid with minimal memory requirements, more specifically exactly one index per grid cell and exactly one index per object reference, and an algorithm for building that data structure in linear time. The hashed grid method reduces memory requirements even further, by using perfect hashing based on row displacement compression. We show that these methods are more efficient in both time and space than traditional methods based on linked lists and dynamic arrays. We also present a more robust grid traversal algorithm. We show that, for applications where time to image or memory usage is important, such as interactive ray tracing and rendering large models, the grid acceleration structure is an attractive alternative.     

Temporal Upsampling of Point Cloud Sequences by Optimal Transport for Plant Growth Visualization

Plant growth visualization from a series of 3D scanner measurements is a challenging task. Time intervals between successive measurements are typically too large to allow a smooth animation of the growth process. Therefore, obtaining a smooth animation of the plant growth process requires a temporal upsampling of the point cloud sequence in order to obtain approximations of the intermediate states between successive measurements. Additionally, there are suddenly arising structural changes due to the occurrence of new plant parts such as new branches or leaves. We present a novel method that addresses these challenges via semantic segmentation and the generation of a segment hierarchy per scan, the matching of the hierarchical representations of successive scans and the segment-wise computation of optimal transport. The transport problems' solutions yield the information required for a realistic temporal upsampling, which is generated in real time. Thereby, our method does not require shape templates, good correspondences or huge databases of examples. Newly grown and decayed parts of the plant are detected as unmatched segments and are handled by identifying corresponding bifurcation points and introducing virtual segments in the previous, respectively successive time step. Our method allows the generation of realistic upsampled growth animations with moderate computational effort.     

Image-Based Network Rendering of Large Meshes for Cloud Computing

Recent advances in sensing and software technologies enable us to obtain large-scale, yet fine 3D mesh models of cultural assets. However, such large models cannot be displayed interactively on consumer computers because of the performance limitation of the hardware. Cloud computing technology is a solution that can process a very large amount of information without adding to each client user’s processing cost. In this paper, we propose an interactive rendering system for large 3D mesh models, stored on a remote environment through a network of relatively small capacity machines, based on the cloud computing concept. Our system uses both model- and image-based rendering methods for efficient load balance between a server and clients. On the server, the 3D models are rendered by the model-based method using a hierarchical data structure with Level of Detail (LOD). On the client, an arbitrary view is constructed by using a novel image-based method, referred to as the Grid-Lumigraph, which blends colors from sampling images received from the server. The resulting rendering system can efficiently render any image in real time. We implemented the system and evaluated the rendering and data transferring performance.     

GPU-based Fast Ray Casting for a Large Number of Metaballs

Metaballs are implicit surfaces widely used to model curved objects, represented by the isosurface of a density field defined by a set of points. Recently, the results of particle‐based simulations have been often visualized using a large number of metaballs, however, such visualizations have high rendering costs. In this paper we propose a fast technique for rendering metaballs on the GPU. Instead of using polygonization, the isosurface is directly evaluated in a per‐pixel manner. For such evaluation, all metaballs contributing to the isosurface need to be extracted along each viewing ray, on the limited memory of GPUs. We handle this by keeping a list of metaballs contributing to the isosurface and efficiently update it. Our method neither requires expensive precomputation nor acceleration data structures often used in existing ray tracing techniques. With several optimizations, we can display a large number of moving metaballs quickly.     

Fast Out-of-Core Octree Generation for Massive Point Clouds

We propose an efficient out-of-core octree generation method for arbitrarily large point clouds. It utilizes a hierarchical counting sort to quickly split the point cloud into small chunks, which are then processed in parallel. Levels of detail are generated by subsampling the full data set bottom up using one of multiple exchangeable sampling strategies. We introduce a fast hierarchical approximate blue-noise strategy and compare it to a uniform random sampling strategy. The throughput, including out-of-core access to disk, generating the octree, and writing the final result to disk, is about an order of magnitude faster than the state of the art, and reaches up to around 6 million points per second for the blue-noise approach and up to around 9 million points per second for the uniform random approach on modern SSDs.     

Interactive Ray Tracing of Large Models Using Voxel Hierarchies

We propose an efficient approach for interactive visualization of massive models with CPU ray tracing. A voxel‐based hierarchical level‐of‐detail (LOD) framework is employed to minimize rendering time and required system memory. In a pre‐processing phase, a compressed out‐of‐core data structure is constructed, which contains the original primitives of the model and the LOD voxels, organized into a kd‐tree. During rendering, data is loaded asynchronously to ensure a smooth inspection of the model regardless of the available I/O bandwidth. With our technique, we are able to explore data sets consisting of hundreds of millions of triangles in real‐time on a desktop PC with a quad‐core CPU.     

Out-of-core selection and editing of huge point clouds

In this paper we present an out-of-core editing system for point clouds, which allows selecting and modifying arbitrary parts of a huge point cloud interactively. We can use the selections to segment the point cloud, to delete points, or to render a preview of the model without the points in the selections. Furthermore, we allow for inserting points into an already existing point cloud. All operations are conducted on a rendering optimized data structure that uses the raw point cloud from a laser scanner, and no additionally created points are needed for an efficient level-of-detail (LOD) representation using this data structure. We also propose an algorithm to alleviate the artifacts when rendering a point cloud with large discrepancies in density in different areas by estimating point sizes heuristically. These estimated point sizes can be used to mimic a closed surface on the raw point cloud, also when the point cloud is composed of several raw laser scans.     

Adaptive Ray‐bundle Tracing with Memory Usage Prediction: Efficient Global Illumination in Large Scenes

This paper proposes an adaptive rendering technique for ray‐bundle tracing. Ray‐bundle tracing can be done by per‐pixel linked‐list construction on a GPU rasterization pipeline. This rasterization based approach offers significant benefits for the efficient generation of light maps (e.g., hardware acceleration, tessellation, and recycling of shaders used in real‐time graphics). However, it is inapplicable to large and complex scenes due to the limited capacity of the GPU memory because it requires a high‐resolution frame buffer and high‐capacity node buffer for the linked‐lists. In addition, memory overflow can potentially occur on the per‐pixel linked‐list since the memory usage of the lists is usually unknown before the rendering process. We introduce an adaptive tiling technique with memory usage prediction. Our method uses an appropriately tiled frame buffer, thus eliminating almost all of the overflow risks thanks to our adaptive tile subdivision scheme. Using this technique, we are able to render high‐quality light maps of large and complex scenes which cannot be computed using previous ray‐bundle based methods.     

Separating semantics from rendering: a scene graph based architecture for graphics applications

A large number of rendering and graphics applications developed in research and industry are based on scene graphs. Traditionally, scene graphs encapsulate the hierarchical structure of a complete 3D scene, and combine both semantic and rendering aspects. In this paper, we propose a clean separation of the semantic and rendering parts of the scene graph. This leads to a generally applicable architecture for graphics applications that is loosely based on the well-known Model-View-Controller (MVC) design pattern for separating the user interface and computation parts of an application. We explore the benefits of this new design for various rendering and modeling tasks, such as rendering dynamic scenes, out-of-core rendering of large scenes, generation of geometry for trees and vegetation, and multi-view rendering. Finally, we show some of the implementation details that have been solved in the process of using this software architecture in a large framework for rapid development of visualization and rendering applications.     

基于CPU-GPU混合加速的SPH流体仿真方法

基于光滑粒子流体力学SPH的流体仿真是虚拟现实技术的重要研究内容,但SPH流体仿真需要大量的计算资源,采用一般计算方法难以实现流体仿真的实时性。流体仿真通常由物理计算、碰撞检测和渲染等部分组成,借助GPU并行加速粒子的物理属性计算和碰撞过程使SPH方法的实时流体仿真成为可能。为了满足流体仿真应用中的真实性和实时性需求,提出一种基于CPU GPU混合加速的SPH流体仿真方法,流体计算部分采用GPU并行加速,流体渲染部分采用基于CPU的OpenMP加速。实验结果表明,基于CPU GPU混合加速的SPH流体仿真方法与CPU实现相比,能显著地减少流体仿真单帧计算时间且能更快速地完成渲染任务。     

高精度高速度曲面建模的3维地表模拟和漫游算法

三维真实感地形场景是生态系统模拟、虚拟地理环境和三维GIS不可或缺的要素,而高精度高速度曲面建模方法可以动态实时构建高精度曲面模型。论文引入高精度高速度曲面建模方法,提出了一种可以动态生成地表模型并实时绘制和漫游三维真实感地表的方法。该方法实时动态建立高精度DEM,结合改进的视点依赖多分辨率绘制方法,随视点改变动态重建并绘制当前DEM。首先建立当前可见场景的DEM,然后通过依赖于视点的简化等处理建立层次细节模型,进行实时场景绘制和漫游,随视点改变,在当前场景基础上动态加点和减点更新DEM及层次细节模型,生成当前可见场景,以次循环。实例测试表明,算法可产生真实感较强的场景,实时漫游效率较高,模拟精确。