ProjetoAM/readdata.m at master · dnr2/ProjetoAM · GitHub

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
function data = readdata(filename)

% Change to the data sets directory
cd datasets;

switch filename
    case 'att_faces'
        data = readattfaces;
    case 'mnistdigits'
        data = readmnistdigits;
    case 'binaryalphabet'
        data = readbinaryalphabet;
    case 'coil20'
        data = readcoil20;
    otherwise
        data = [];
        display('Invalid file name!')
end

% Change back to the original directory
cd ..;

end

% READING FUNCTIONS %

function data = readattfaces

% Preallocate for performance
% Since each image is a 23x28 image (28x23 matrix)
% data is a 400x644 matrix (400 images, 644 = 23*28 features)
data = zeros(400, 644);

% Change directory
cd att_faces;
numfolders = 40;
numimages = 10;

% For each folder
for i = 1:numfolders
    foldername = sprintf('s%d', i);
    cd (foldername);
    for j = 1:numimages
        imagename = sprintf('%d.pgm', j);
        imagematrix = imread(imagename);
        data((i - 1)*numimages + j, :) = reshape(imagematrix', 1, numel(imagematrix));
    end
    cd ..;
end

% Back to the original directory
cd ..;
end

function data = readmnistdigits

% Load the data set
% test{0-9} test sets
% train{0-9} train sets
load mnist_all.mat
numimages = 50;

% Since each image is a 28x28 image
% data is a 500x784 matrix (500 = 10*50 images, 784 = 28*28 features)
% Read the images into the data matrix
data = [train0(1:numimages, :); train1(1:numimages, :);
    train2(1:numimages, :); train3(1:numimages, :);
    train4(1:numimages, :); train5(1:numimages, :);
    train6(1:numimages, :); train7(1:numimages, :);
    train8(1:numimages, :); train9(1:numimages, :)];

% Convert it to a matrix of doubles
data = double(data);

end

function data = readbinaryalphabet

% Preallocate for performance
% Since each image is a 16x20 binary image (20x16 matrix)
% data is a 1014x320 matrix (1014 = 26*39 images, 320 = 16*20 features)
data = zeros(1014, 320);

% Load the data set
% classcounts 1x36 cell
% classlabels 1x36 cell
% dat 36x39 cell
% numclass = 36
%
load binaryalphadigs.mat

% Since we want only the alphabet digits
% We read only from dat{11, :} until dat{36, :}
numimages = classcounts{1}; % All the classes have the same number of examples

% Read the images into the data matrix
for i = 11:numclass
    for j = 1:numimages
        imagematrix = dat{i, j};
        data((i - 11)*numimages + j, :) = reshape(imagematrix', 1, numel(imagematrix));
    end
end

end

function data = readcoil20

% Preallocate for performance
% Since each image is a 32x32 image
% data is a 1000x1024 matrix (1000 images, 1024 = 32*32 features)
data = zeros(1000, 1024);

% Change directory
cd coil20;

% Number of objects and images for each object
numobjects = 20;
numimages = 50;

% Read the images into the data matrix
for i = 1:numobjects
    for j = 0:(numimages - 1)
        imagename = sprintf('obj%d__%d.png', i, j);
        imagematrix = imread(imagename);
        data((i - 1)*numimages + (j + 1), :) = reshape(imagematrix', 1, numel(imagematrix));
    end
end

% Back to the original directory
cd ..;
end