Paper Info
Title
Statistical Modeling of a Moving-Image Sequence in the 3-D DFT Domain
Abstract
For the statistical modeling of DFT coefficients of random signals, this paper presents a multidimensional 2-Component Spherically-Symmetric Gaussian Mixture (2-C S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> GM) distribution model based on a degradation model and signal’s sparsity, and constructs an iterative solver providing a unique solution of its model-parameter estimation problem along the lines of the moment approach. Moreover, this paper experimentally evaluates statistical accuracy of our model-parameter estimation method with the iterative solver. Furthermore, applying our model-parameter estimation method to sample sequences of 3-D DFT coefficients of a moving-image sequence, this paper shows that the model fitting with the 2-C S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> GM distribution model provides an estimate of signal sparsity for the moving-image sequence. Lastly, applying the 2-C S <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</sup> GM statistical modeling to moving-image denoising, this paper shows that the modeling is a potential tool for moving-image restoration in the 3-D DFT domain.
Year
DOI
Venue
2018
10.1109/TENCON.2018.8650356
TENCON 2018 - 2018 IEEE Region 10 Conference
Keywords
Field
DocType
Solid modeling,Discrete Fourier transforms,Estimation,Iterative methods,Gaussian distribution,Probability distribution,Computational modeling
Noise reduction,Iterative method,Computer science,Algorithm,Electronic engineering,Probability distribution,Gaussian,Statistical model,Solid modeling,Solver,Image sequence
Conference
ISSN
ISBN
Citations 
2159-3442
978-1-5386-5457-6
0
PageRank 
References 
Authors
0.34
0
2
Name
Order
Citations
PageRank
Takahiro Saito110030.46
Takashi Komatsu211333.96