-
Notifications
You must be signed in to change notification settings - Fork 61
/
linop_TV3D.m
197 lines (177 loc) · 6.62 KB
/
linop_TV3D.m
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
function op = linop_TV3D( sz, variation, action )
%LINOP_TV3D 3D Total-Variation (TV) linear operator.
% OP = LINOP_TV3D( SZ ) returns a handle to a TFOCS linear operator that
% implements the total variation linear operator on an M x N x P grid;
% that is, to be applied to volume stacks of size [M,N,P].
% By default, it expects to operate on M*N*P x 1 vectors
% but if SZ = {M,N,P}, then expects to operate on M x N x P matrices
%
% TV = LINOP_TV3D( X ) returns ||X||_TV if X is bigger than 2 x 2 x 2
%
% OP = LINOP_TV3D( SZ, VARIANT )
% if VARIANT is 'regular' (default),
% ... TODO
% if VARIANT is 'circular',
% ... TODO
%
% [...] = LINOP_TV3D(SZ, VARIATION, ACTION )
% if ACTION is 'handle', returns a TFOCS function handle (default)
% if ACTION is 'cvx', returns a function handle suitable for CVX
% if ACTION is 'matrix', returns the explicit TV matrix
% (real part corresponds to horizontal differences,
% imaginary part correspond to vertical differences)
% if ACTION is 'norm', returns an estimate of the norm
%
% Contributed by Mahdi Hosseini ([email protected])
% Has not been extensively tested
error(nargchk(1,3,nargin));
if nargin < 2 || isempty(variation), variation = 'regular'; end
if nargin < 3 || isempty(action), action = 'handle'; end
CALCULATE_TV = false;
if numel(sz) > 6
CALCULATE_TV = true;
X = sz;
sz = size(X);
end
nDim = numel(sz);
if iscell(sz)
n1 = sz{1};
n2 = sz{2};
n3 = sz{3};
else
n1 = sz(1);
n2 = sz(2);
n3 = sz(3);
end
% Setup the Total-Variation operators
mat = @(x) reshape(x,n1,n2,n3);
if strcmpi(action,'matrix') || strcmpi(action,'cvx')
I1 = eye(n1);
I2 = eye(n2);
I3 = eye(n3);
switch lower(variation)
case 'regular'
e = ones(max([n1,n2,n3]),1);
e2 = e;
e2(n2:end) = 0;
J = spdiags([-e2,e], 0:1,n2,n2);
Dh = kron(I3,kron(J,I1)); % horizontal differences, sparse matrix
% see also blkdiag
e2 = e;
e2(n1:end) = 0;
J = spdiags([-e2,e], 0:1,n1,n1);
Dv = kron(I3,kron(I2,J)); % vertical differences, sparse matrix
e2 = e;
e2(n3:end) = 0;
J = spdiags([-e2,e], 0:1,n3,n3);
Dd = kron(J,kron(I2,I1)); % Depth differences, sparse matrix
case 'circular'
e = ones(max([n1,n2,n3]),1);
e2 = e;
% e2(n2:end) = 0;
J = spdiags([-e2,e], 0:1,n2,n2);
J(end,1) = 1;
Dh = kron(I3, kron(J,I1)); % horizontal differences, sparse matrix
% see also blkdiag
e2 = e;
% e2(n1:end) = 0;
J = spdiags([-e2,e], 0:1,n1,n1);
J(end,1) = 1;
Dv = kron(I3, kron(I2,J)); % vertical differences, sparse matrix
e2 = e;
% e2(n1:end) = 0;
J = spdiags([-e2,e], 0:1,n3,n3);
J(end,1) = 1;
Dd = kron(J, kron(I2,I1)); % vertical differences, sparse matrix
end
if strcmpi(action,'matrix')
op = [Dh;Dv;Dd];
else
% "norms" is a CVX function
op = @(X) sum( norms( [Dh*X(:), Dv*X(:), Dd*X(:)]' ) );
end
return;
end
switch lower(variation)
case 'regular'
Dh = @(X) vec( [diff(X,1,2), zeros(n1,1,n3)] );
Dv = @(X) vec( [diff(X,1,1); zeros(1,n2,n3)] );
Dd = @(X) [vec(diff(X,1,3)); vec(zeros(n1,n2,1))];
diff_h = @(X) [zeros(n1,1,n3),X(:,1:end-1,:)] - ...
[X(:,1:end-1,:),zeros(n1,1,n3)];
diff_v = @(X) [zeros(1,n2,n3);X(1:end-1,:,:)] - ...
[X(1:end-1,:,:);zeros(1,n2,n3)];
diff_d = @(X) mat([vec(zeros(n1,n2,1));vec(X(:,:,1:end-1))] - ...
[vec(X(:,:,1:end-1));vec(zeros(n1,n2,1))]);
case 'circular'
% For circular version, 2 x 2 case is special.
% error('not yet implemented');
Dh = @(X) vec( [diff(X,1,2), X(:,1,:) - X(:,end,:)] );
Dv = @(X) vec( [diff(X,1,1); X(1,:,:) - X(end,:,:)] );
Dd = @(X) [vec(diff(X,1,3)); vec(X(:,:,1) - X(:,:,end))];
% diff_h needs to be checked
diff_h = @(X) [X(:,end,:),X(:,1:end-1,:)] - X;
% diff_v needs to be checked
diff_v = @(X) [X(end,:,:);X(1:end-1,:,:)] - X;
% diff_d needs to be checked
diff_d = @(X) mat([vec(X(:,:,end));vec(X(:,:,1:end-1))]) - X;
otherwise
error('Bad variation parameter');
end
if iscell(sz)
Dh_transpose = @(X) diff_h(mat(X)) ;
Dv_transpose = @(X) diff_v(mat(X)) ;
Dd_transpose = @(X) diff_d(mat(X)) ;
else
Dh_transpose = @(X) vec( diff_h(mat(X)) );
Dv_transpose = @(X) vec( diff_v(mat(X)) );
Dd_transpose = @(X) vec( diff_d(mat(X)) );
end
%% TV & TVt Definitions
TV = @(x) [Dh(mat(x)); Dv(mat(x)); Dd(mat(x))];
firstThird = @(x) x(1: n1*n2*n3);
secondThird= @(x) x(n1*n2*n3+1: 2*n1*n2*n3);
thirdThird= @(x) x(2*n1*n2*n3+1: 3*n1*n2*n3);
TVt = @(z) ( Dh_transpose(firstThird(z)) +...
Dv_transpose(secondThird(z)) +...
Dd_transpose(thirdThird(z)));
%%
if CALCULATE_TV
op = norm( TV(X), 1 );
return;
end
if strcmpi(action,'norm')
% to compute max eigenvalue, I use a vector
% that is very likely to be the max eigenvector:
% matrix with every entry alternating -1 and 1
even = @(n) ~( n - 2*round(n/2) ); % returns 1 if even, 0 if odd
Y = zeros( n1 + even(n1), n2 + even(n2), n3 + even(n3) );
nn = numel(Y);
Y(:) = (-1).^(1:nn);
Y = Y(1:n1,1:n2,1:n3);
op = norm( TV(Y) )/norm(Y(:));
% Nearly equivalent to:
% norm(full( [real(tv); imag(tv)] ) )
% where tv is the matrix form
else
if iscell(sz)
szW = { [n1,n2,n3], [n1*n2*n3,1] };
else
szW = prod(sz); % n1 * n2
szW = [nDim*szW, nDim*szW];
end
op = @(x,mode)linop_tv_r2c(szW,TV,TVt,x,mode);
end
function y = linop_tv_r2c( sz, TV, TVt, x, mode )
switch mode,
case 0, y = sz;
case 1, y = TV( realcheck( x ) );
case 2, y = realcheck( TVt( x ) );
end
function y = realcheck( y )
if ~isreal( y ),
error( 'Unexpected complex value in linear operation.' );
end
% TFOCS v1.3 by Stephen Becker, Emmanuel Candes, and Michael Grant.
% Copyright 2013 California Institute of Technology and CVX Research.
% See the file LICENSE for full license information.