Denoising For Surface Reconstruction

MPhil Thesis Defence


Title: "Denoising For Surface Reconstruction"

By

Mr. Man-Kit Lau


Abstract

We present an algorithm to denoise an unorganized point cloud which 
contains noise, white noise and outliers for surface reconstruction with 
the assumption that the points close to the real surface are uniformly 
distributed. Our algorithm first remove the outliers, white noise and high 
noise points base on the point cloud behavior and simple statistical 
analysis. After that, our algorithm denoise the remaining points by a 
modified Laplacian Smoothing algorithm while our algorithm works for 
unorganized point cloud. Finally, we apply Robust Cocone algorithm to 
reconstruct the surface from the denoised point cloud followed by some 
postprocessing on the reconstructed surface for constructing the 
boundaries and further smoothing. Our algorithm applies a variant of the 
standard octree structure to manipulate the points and this makes our 
algorithm more efficient. The experimental results show that our algorithm 
can generate smooth surface even though the noise level are very large and 
the running time is fast. The data sets that we have experimented includes 
some raw data with contaminated by artificial noise, white noise and 
cluster of outliers. For the comparisons, the reconstructed surface of our 
algorithm is compared with the reconstructed surface of applying Robust 
Cocone algorithm on the point cloud without denoising in order to show the 
effectiveness of our algorithm. We also compare our algorithm with the 
Adaptive Moving Least-Squares (AMLS) algorithm in terms of running time 
and quality. The comparison results shows that our algorithm can improve 
the surface reconstructed from Robust Cocone algorithm, and is comparable 
with the AMLS algorithm while our algorithm runs much faster than AMLS 
algorithm.


Date:				Friday, 24 August 2012

Time:				2:00pm – 4:00pm

Venue:				Room 5506
 				Lifts 25/26

Committee Members:		Prof. Siu-Wing Cheng (Supervisor)
 				Dr. Ke Yi (Chairperson)
 				Dr. Chiew-Lan Tai


**** ALL are Welcome ****