de
en
Schliessen
Detailsuche
Bibliotheken
Projekt
Impressum
Datenschutz
zum Inhalt
Detailsuche
Schnellsuche:
OK
Ergebnisliste
Titel
Titel
Inhalt
Inhalt
Seite
Seite
Im Dokument suchen
Mensmann, Jörg ; Mensmann, J.: Exploiting spatial and temporal coherence in GPU-based volume rendering. 2010
Inhalt
Preface
Introduction
General-Purpose Programming on Graphics Processors
GPGPU Approaches
Stream Processing
History of stream processing in graphics
CUDA Basics
Differences between CUDA and graphics programming
Concepts of Volume Rendering
Theoretical Background
Volume Rendering Approaches
Direct volume rendering
Transfer functions
Illumination models
Performance and image quality
GPU-based Raycasting
Accelerating Volume Raycasting
Empty space skipping
Early ray termination
Bricking and multi-resolution approaches
Voreen: The Volume Rendering Engine
Data-flow Concept
User Interface
Integration of GPU-based Raycasting
Adding New Components
Performance Considerations
Empty Space Skipping using Occlusion Frustums
Related Work
Impact of Hardware Restrictions on GPU-based Raycasting
Optimizing the Proxy Geometry for Space Leaping
Occlusion frustums as proxy geometry
Clipping the occlusion volume
Possible extensions to the occlusion frustum approach
GPU Implementation
Analyzing first-hit points
Generating occlusion frustums
Integration into Voreen
Data-flow network
Implementing the optimized proxy geometry processor
Results
Performance evaluation
Discussion
Summary
Applying GPU Stream Processing to Volume Raycasting
Related Work
GPU-based volume raycasting
GPU stream processing
Raycasting with CUDA
Using the CUDA architecture for raycasting
3D texture caching
Accelerating raycasting
Implementing Basic Raycasting
Fragment shader implementation
CUDA implementation
Slab-based Raycasting
Slab-based approach
CUDA implementation
Integration into Voreen
Volume handling
Network integration
Results
Testing methodology
Basic raycaster
Slab-based raycaster
Discussion
Summary
Lossless Compression for Rendering Time-Varying Volume Data
Related Work
Hybrid Compression Scheme
Data properties and hardware limitations
Two-stage compression approach
Subdividing the volume into bricks for compression
Main compression algorithm
Prediction schemes
Variable-length coding
Data preprocessing
GPU-supported Decompression Pipeline
Multi-threaded loading and LZO decompression
Asynchronous data transfer to the GPU
Decoding and brick assembly
Rendering
Integration into Voreen
Preprocessing and on-disk storage format
Network integration
Results
Test data sets
Compression ratio
Rendering speed
Summary
Conclusions
Source Code
Fragment Shader for Volume Raycasting
CUDA Kernel for Volume Raycasting
Index of Data Sets
Bibliography
Acronyms