mirror of
https://github.com/jlengrand/LAFF.git
synced 2026-03-10 08:31:21 +00:00
57 lines
1.4 KiB
Matlab
Executable File
57 lines
1.4 KiB
Matlab
Executable File
% Close all figures
|
|
close all
|
|
|
|
% Read the picture in building.png into matrix B
|
|
[ B, map ] = imread( 'building.png' );
|
|
|
|
% Show the picture in a figure
|
|
imshow( B );
|
|
|
|
% Choose the (approximate) number of columns you want to use for the
|
|
% aproximation.
|
|
k = 30;
|
|
|
|
% Extract the number of rows and columns in B
|
|
[ m, n ] = size( B )
|
|
|
|
% B starts as an integer array. To do the computations, we need to
|
|
% make it into floating point numbers.
|
|
|
|
B = double( B );
|
|
|
|
% Pick A as the matrix consisting of about k equally spaced columns of
|
|
% B.
|
|
stride = round( n / k );
|
|
A = B( :, 1:stride:n );
|
|
|
|
% Replace the comments below with their respective operations from the notebook
|
|
|
|
% C = A^T A
|
|
C = A' * A;
|
|
|
|
% V = A^T B
|
|
V = A' * B;
|
|
|
|
% Overwrite C with its LU factorization
|
|
C = LU_unb_var5( C );
|
|
|
|
% Extract the unit lower triangular matrix L and upper triangular matrix U.
|
|
L = tril( C, -1 ) + eye( size( C ) );
|
|
U = triu( C );
|
|
|
|
% Solve L(UX) = V, overwriting V with X
|
|
V = L \ V;
|
|
V = U \ V;
|
|
|
|
% Create a new figure in which you show the approximated picture. Notice
|
|
% that A and V are matrices of doubles, which much be converted to unsigned
|
|
% 8 byte integers for imshow to work.
|
|
figure
|
|
imshow( uint8( A * V ) );
|
|
|
|
% The two figures may be on top of each other. You may have to move the
|
|
% second one next to the first one.
|
|
|
|
% Play around by changing the number of columns you use for the
|
|
% approximation!
|