Efficient on-line visualization of 3D textured models is essential for a variety of applications including not only video games and e-commerce, but also virtual museums and Tele-health. To visualize 3D objects online, it is necessary to quickly adapt both mesh and texture to the available computational or network resources. Early research has shown that, after reaching a minimum required mesh density, high-resolution texture has more impact on human perception than a denser mesh. Given limited bandwidth, important issues include how to extract features that best represent the original object, and how to allocate resources between mesh and texture data to achieve optimal perceptual quality. In this thesis, I propose a textured mesh (TexMesh) framework, which applies scale-space filtering (SSF) and perceptual evaluation to extract 3D features for textured mesh simplification and transmission. The appropriate level-of-detail (LOD) is automatically selected, based on the object size on the display device. Mesh refinement is guided by the just-noticeable-difference (JND) threshold, below which redundant mesh data are suppressed. Weber's law was applied to locate the JND, and perceptual experimental results showed that the threshold for 3D TexMesh is around 0.10, which is consistent with other psycho-visual experimental results on Weber's law in the literature. I apply the fragmentation approach on texture transmission to facilitate quality and bandwidth adaptation. Texture quality assignment is based on a visual quality prediction (VQP) model. On-line transmission can be performed efficiently using statistics gathered during preprocessing, which are stored in a priority queue and lookup tables. Quality of Service (QoS) requested by a client site is met by applying an efficient adaptive Harmonic Time Compensation Algorithm (HTCA) to ensure optimal use of the specified time and available bandwidth, while preserving satisfactory quality.; This thesis presents a new approach integrating feature extraction, mesh simplification, texture reduction, bandwidth adaptation, and perceptual evaluation into a multi-scale visualization framework.
展开▼
