Multiresolution Model Compression Using 3-D Wavelets

Three-dimensional (3-D) objects are often represented by geometric models in applications dealing with virtual reality, augmented reality, and cyberspace. Surface representations can provide an effective visualization of these objects. Polygonal models are the most prevalent type of surface representation. Recently, multiresolution representation (surface simplification) of polygonal models has been proposed to meet the requirements of easy manipulation, progressive transmission, effective visualization, and economical storage. Based on 3-D wavelet transforms, we propose a framework for multiresolution modeling, in which a surface in 3-D space is treated as an extension of an edge in 2-D space. In addition, we utilize a volumetric surface model which can be compressed simultaneously at multiple levels of detail (LODs). The compression efficiency of this framework is further improved by applying a lattice vector quantization and an arithmetic coding technique on the compact wavelet coefficients. This modeling framework can be easily integrated with object models obtained from volumetric implicit-surface representations of multiple range images or with models from volume data. Models constructed from real objects and synthetic objects have been used to demonstrate the compression efficiency of this wavelet-based framework.