A Survey on Surface Denoising Methods

PhD Qualifying Examination


Title: "A Survey on Surface Denoising Methods"

Mr. Youyi Zheng


Abstract:

Fundamental principles of geometry processing give rise to a wide range of 
useful tools for manipulating and transforming surfaces. Among which, a 
requirement of the underlying surface being smooth is usually a prior. 
However, raw data acquired through various resources inevitably contain 
noise and must be denoised before further processing. Eliminating the 
noise while recovering the underlying surface geometry as much as possible 
remains a key challenge of surface denoising. In recent years, 
feature-preserving surface denoising techniques have emerged as powerful 
tools. A variety of non-linear filters, including diffusion-based 
anisotropic filters, bilateral filters, non-local means filters and 
various ad-hoc normal filters designed for specific applications have been 
introduced. Despite the demonstrated good properties and the solid 
theoretical background, there lacks a comprehensive study on the 
performances and limitations of existing denoising techniques as well as 
their inter-relations. This survey provides an overview of the 
state-of-art surface denoising methods. We analyze and categorize the 
previous denoising solutions into various classes based on the theoretical 
models used, the algorithmic techniques employed and the domains focused 
on. To further investigate these approaches, several representative 
techniques were implemented and their results compared.


Date:     		Tuesday, 2 June 2009

Time:                   4:00pm-6:00pm

Venue:                  Room 5486
 			lifts 25-26

Committee Members:      Dr. Chiew-Lan Tai (Supervisor)
 			Dr. Pedro Sander (Chairperson)
 			Prof. Long Quan
 			Dr. Chi-Keung Tang


**** ALL are Welcome ****