# 9.1 A class of fast algorithms for total variation image restoration  (Page 4/6)

 Page 4 / 6
$\begin{array}{c}\hfill \underset{\mathbf{w},u}{min}\sum _{i}\parallel {\mathbf{w}}_{i}\parallel +\frac{\beta }{2}\sum _{i}\parallel {\mathbf{w}}_{i}-{D}_{i}{u\parallel }^{2}+\frac{\mu }{2}{\parallel Ku-f\parallel }^{2},\end{array}$

where $\beta \gg 0$ is a penalty parameter. It is well known that the solution of ( ) converges to that of ( ) as $\beta \to \infty$ . In the following, we concentrate on problem ( ).

## Basic algorithm

The benefit of ( ) is that while either one of the two variables $u$ and $\mathbf{w}$ is fixed, minimizing the objective function with respect to the other has a closed-form formula that we willspecify below. First, for a fixed $u$ , the first two terms in ( ) are separable with respect to ${\mathbf{w}}_{i}$ , and thus the minimization for $\mathbf{w}$ is equivalent to solving

$\begin{array}{c}\hfill \underset{{\mathbf{w}}_{i}}{min}\parallel {\mathbf{w}}_{i}\parallel +\frac{\beta }{2}{\parallel {\mathbf{w}}_{i}-{D}_{i}u\parallel }^{2},\phantom{\rule{1.em}{0ex}}i=1,2,...,{n}^{2}.\end{array}$

It is easy to verify that the unique solutions of ( ) are

$\begin{array}{c}\hfill {\mathbf{w}}_{i}=max\left\{\parallel ,{D}_{i},u\parallel -,\frac{1}{\beta },,,0\right\}\frac{{D}_{i}u}{\parallel {D}_{i}u\parallel },\phantom{\rule{1.em}{0ex}}i=1,...,{n}^{2},\end{array}$

where the convention $0·\left(0/0\right)=0$ is followed. On the other hand, for a fixed $\mathbf{w}$ , ( ) is quadratic in $u$ and the minimizer $u$ is given by the normal equations

$\begin{array}{c}\hfill \left(\sum _{i},{D}_{i}^{\top },{D}_{i},+,\frac{\mu }{\beta },{K}^{\top },K\right)u=\sum _{i}{D}_{i}^{\top }{\mathbf{w}}_{i}+\frac{\mu }{\beta }{K}^{\top }f.\end{array}$

By noting the relation between $D$ and ${D}_{i}$ and a reordering of variables, ( ) can be rewritten as

$\begin{array}{c}\hfill \left({D}^{\top },D,+,\frac{\mu }{\beta },{K}^{\top },K\right)u={D}^{\top }w+\frac{\mu }{\beta }{K}^{\top }f,\end{array}$

where

$\begin{array}{c}\hfill w\triangleq \left(\begin{array}{c}{w}_{1}\\ {w}_{2}\end{array}\right)\in {\mathbb{R}}^{2{n}^{2}}\phantom{\rule{0.277778em}{0ex}}\mathrm{and}\phantom{\rule{0.277778em}{0ex}}{w}_{j}\triangleq \left(\begin{array}{c}{\left({\mathbf{w}}_{1}\right)}_{j}\\ ⋮\\ {\left({\mathbf{w}}_{{n}^{2}}\right)}_{j}\end{array}\right),\phantom{\rule{0.277778em}{0ex}}j=1,2.\end{array}$

The normal equation ( ) can also be solved easily provided that proper boundary conditions are assumed on $u$ . Since both the finite difference operations and the convolution are notwell-defined on the boundary of $u$ , certain boundary assumptions are needed when solving ( ). Under the periodic boundary conditions for $u$ , i.e. the 2D image $u$ is treated as a periodic function in both horizontal and vertical directions, ${D}^{\left(1\right)}$ , ${D}^{\left(2\right)}$ and $K$ are all block circulant matrices with circulant blocks; see e.g. , . Therefore, the Hessianmatrix on the left-hand side of ( ) has a block circulant structure and thus can be diagonalized by the 2D discreteFourier transform $\mathbf{F}$ , see e.g. . Using the convolution theorem of Fourier transforms, the solution of( ) is given by

$\begin{array}{c}\hfill u={\mathbf{F}}^{-1}\left(\frac{\mathbf{F}\left({D}^{\top },w,+,\left(\mu /\beta \right),{K}^{\top },f\right)}{\mathrm{diag}\left(\mathcal{F},\left(,{D}^{\top },D,+,\left(\mu /\beta \right),{K}^{\top },K,\right)\right)}\right),\end{array}$

where the division is implemented by componentwise. Since all quantities but $w$ are constant for given $\beta$ , computing $u$ from ( ) involves merely the finite difference operation on $w$ and two FFTs (including one inverse FFT), once the constant quantities are computed.

Since minimizing the objective function in ( ) with respect to either $\mathbf{w}$ or $u$ is computationally inexpensive, we solve ( ) for a fixed $\beta$ by an alternating minimization scheme given below.

Algorithm :

• Input $f$ , $K$ and $\mu >0$ . Given $\beta >0$ and initialize $u=f$ .
• While“not converged”, Do
• Compute $\mathbf{w}$ according to ( ) for fixed $u$ .
• Compute $u$ according to ( ) for fixed $w$ (or equivalently $\mathbf{w}$ ).
• End Do

## Optimality conditions and convergence results

To present the convergence results of Algorithm "Basic Algorithm" for a fixed $\beta$ , we make the following weak assumption.

Assumption 1 $\mathcal{N}\left(K\right)\cap \mathcal{N}\left(D\right)=\left\{0\right\}$ , where $\mathcal{N}\left(·\right)$ represents the null space of a matrix.

Define

$\begin{array}{c}\hfill M={D}^{\top }D+\frac{\mu }{\beta }{K}^{\top }K\phantom{\rule{0.277778em}{0ex}}\phantom{\rule{0.277778em}{0ex}}\phantom{\rule{3.33333pt}{0ex}}\mathrm{and}\phantom{\rule{3.33333pt}{0ex}}\phantom{\rule{0.277778em}{0ex}}\phantom{\rule{0.277778em}{0ex}}T=D{M}^{-1}{D}^{\top }.\end{array}$

Furthermore, we will make use of the following two index sets:

$\begin{array}{c}\hfill L=\left\{i\phantom{\rule{3.33333pt}{0ex}}:\phantom{\rule{3.33333pt}{0ex}}\parallel ,{D}_{i},{u}^{*},\parallel <,\frac{1}{\beta }\right\}\phantom{\rule{4.pt}{0ex}}\phantom{\rule{4.pt}{0ex}}\text{and}\phantom{\rule{4.pt}{0ex}}\phantom{\rule{4.pt}{0ex}}E=\left\{1,...,{n}^{2}\right\}\setminus L.\end{array}$

Under Assumption 1, the proposed algorithm has the following convergence properties.

Theorem 1 For any fixed $\beta >0$ , the sequence $\left\{\left({w}^{k},{u}^{k}\right)\right\}$ generated by Algorithm "Basic Algorithm" from any starting point $\left({w}^{0},{u}^{0}\right)$ converges to a solution $\left({w}^{*},{u}^{*}\right)$ of ( ). Furthermore, the sequence satisfies

• $\parallel {w}_{E}^{k+1}-{w}_{E}^{*}\parallel \le \sqrt{\rho \left({\left({T}^{2}\right)}_{EE}\right)}\parallel {w}_{E}^{k}-{w}_{E}^{*}\parallel ;$
• $\parallel {u}^{k+1}-{u}^{*}{\parallel }_{M}\le \sqrt{\rho \left({T}_{EE}\right)}{\parallel {u}^{k}-{u}^{*}\parallel }_{M};$

anyone know any internet site where one can find nanotechnology papers?
research.net
kanaga
Introduction about quantum dots in nanotechnology
what does nano mean?
nano basically means 10^(-9). nanometer is a unit to measure length.
Bharti
do you think it's worthwhile in the long term to study the effects and possibilities of nanotechnology on viral treatment?
absolutely yes
Daniel
how to know photocatalytic properties of tio2 nanoparticles...what to do now
it is a goid question and i want to know the answer as well
Maciej
Abigail
for teaching engĺish at school how nano technology help us
Anassong
Do somebody tell me a best nano engineering book for beginners?
there is no specific books for beginners but there is book called principle of nanotechnology
NANO
what is fullerene does it is used to make bukky balls
are you nano engineer ?
s.
fullerene is a bucky ball aka Carbon 60 molecule. It was name by the architect Fuller. He design the geodesic dome. it resembles a soccer ball.
Tarell
what is the actual application of fullerenes nowadays?
Damian
That is a great question Damian. best way to answer that question is to Google it. there are hundreds of applications for buck minister fullerenes, from medical to aerospace. you can also find plenty of research papers that will give you great detail on the potential applications of fullerenes.
Tarell
what is the Synthesis, properties,and applications of carbon nano chemistry
Mostly, they use nano carbon for electronics and for materials to be strengthened.
Virgil
is Bucky paper clear?
CYNTHIA
carbon nanotubes has various application in fuel cells membrane, current research on cancer drug,and in electronics MEMS and NEMS etc
NANO
so some one know about replacing silicon atom with phosphorous in semiconductors device?
Yeah, it is a pain to say the least. You basically have to heat the substarte up to around 1000 degrees celcius then pass phosphene gas over top of it, which is explosive and toxic by the way, under very low pressure.
Harper
Do you know which machine is used to that process?
s.
how to fabricate graphene ink ?
for screen printed electrodes ?
SUYASH
What is lattice structure?
of graphene you mean?
Ebrahim
or in general
Ebrahim
in general
s.
Graphene has a hexagonal structure
tahir
On having this app for quite a bit time, Haven't realised there's a chat room in it.
Cied
what is biological synthesis of nanoparticles
what's the easiest and fastest way to the synthesize AgNP?
China
Cied
types of nano material
I start with an easy one. carbon nanotubes woven into a long filament like a string
Porter
many many of nanotubes
Porter
what is the k.e before it land
Yasmin
what is the function of carbon nanotubes?
Cesar
I'm interested in nanotube
Uday
what is nanomaterials​ and their applications of sensors.
Got questions? Join the online conversation and get instant answers!