3D Geometric Coding using Mixture Models and the Estimation Quantization Algorithm

دانشگاه آزاد اسلامی اصفهان(خوراسگان)

منو

رکورد قبلی

رکورد بعدی

" 3D Geometric Coding using Mixture Models and the Estimation Quantization Algorithm "

Document Type	:	Latin Dissertation
Language of Document	:	English
Record Number	:	151830
Doc. No	:	ET23622
Main Entry	:	Sridhar Lavu
Title Proper	:	3D Geometric Coding using Mixture Models and the Estimation Quantization Algorithm
Note	:	This document is digital این مدرک بصورت الکترونیکی می باشد
Abstract	:	3D surfaces are used in applications such as animations, 3D object modeling andvisualization. The geometries of such surfaces are often approximated using polygonalmeshes. This thesis aims to compress 3D geometry meshes by using an algorithmbased on normal meshes and the Estimation-Quantization (EQ) algorithm. Normalmeshes are multilevel representations where finer level vertices lie in a direction normalto the local surface and therefore compress the vertex data to one scalar value pervertex. A mixture distribution model is used for the wavelet coefficients. The EQalgorithm uses the local neighborhood information and Rate-Distortion optimizationto encode the wavelet coefficients. We achieve performance gains of 0.5-ldB comparedto the zerotree coder for normal meshes..........-....,....-...,..tested for theQ1 PC1 bus cardBoth these projects mere sofixare des elopment efforts tonards contributing to dlfferentaspects of Roboucs and lZ1echatronics projects m the Controls and Roboucs Group..
Subject	:	Electericl tess
	:	برق
electronic file name	:	TL46866.pdf
Title and statement of responsibility and	:	3D Geometric Coding using Mixture Models and the Estimation Quantization Algorithm [Thesis]