DynamicSFM/MyImageSegmentation3.m at master · TooCoo/DynamicSFM · 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
132
133
134
135
136
close all; clear all; clc;

nF = 100;


%Load data from my SIFT matcher and F estimator

%Unary = textscan('FWeights', 'f');
fileID = fopen('FWeights');
[A,countF] = fscanf(fileID, '%f');
fclose(fileID);

fileID = fopen('neighbours');
[B,countN] = fscanf(fileID, '%d');
fclose(fileID);

fileID = fopen('neighbourCost');
[C,countNC] = fscanf(fileID, '%f');
fclose(fileID);

fileID = fopen('points1');
[D,countP] = fscanf(fileID, '%f');
fclose(fileID);

nPoints = countP/2;
nNeighbours = countN/nPoints;

unary = reshape(A', nPoints, nF);
neighbourhood = reshape(B', nPoints, nNeighbours); neighbourhood = neighbourhood';
neighbourCost = reshape(C', nPoints, nNeighbours); neighbourCost = neighbourCost';
currentLabel = zeros(1, nPoints);
costOfModelApp = ones(1, nF) * 0.5; costOfModelApp = costOfModelApp';
outlier = 15.0;
points = reshape(D', nPoints, 2);

%neighbourhoodZero = neighbourhood == 0;

neighbourCostMax = max(neighbourCost(:));
%unaryMax = max(unary(:));
%unary = unary * (1/unaryMax);
%unary = unary .* unary;
%unary = unary * 20;

%neighbourhood = neighbourhood + ones(size(neighbourhood)) - neighbourhoodZero;
%neighbourCost = ones(size(neighbourCost))*0.1;

neighbourCost = ones(size(neighbourCost))*2;

%max is around 5k


max_d = 5000;

for i = 1:nPoints
   for j = 1:nNeighbours

       %neighbourCost(j,i) = neighbourCost(j,i);

       if neighbourhood(j,i) == i
           neighbourhood(j, i) = 0;
       end

   end
end


%figure;
%plot(neighbourCost(:,1), '.r');

rng(10);
cols = rand(nF, 3);
%cols = hsv(nF);

figure;
%img = imread('2_1m.jpg');
%img = rgb2gray(imread('6_1m.png'));
img = rgb2gray(imread('blend_2.png'));
%img = imread('myD1m.jpg');
imshow(img);


nDrawn = 0;


for j = 1:1

    [models internal]= expand(unary, neighbourhood, neighbourCost, currentLabel, costOfModelApp, outlier);

    hold on;
    for i = 1:nPoints
        %plot(points(i,1),points(i,2),'r.','MarkerSize',20)
        if(internal(i,1) ~= 0)
            %nDrawn = nDrawn ;
            plot(points(i,1),points(i,2),'r.', 'Color', cols(internal(i, 1), :),'MarkerSize',20);
        end
    end
    hold off;


end

 %lets check that the neighbours are indeed the neighbours
 figure;
 imshow(img);
 for i = 1:nPoints
     imshow(img);
     hold on;

     for j = 1:nNeighbours

         index = neighbourhood(j,i);

         if index ~=0

             x = points(neighbourhood(j,i), 1);
             y = points(neighbourhood(j,i), 2);


             plot(x, y,'r.', 'Color', cols(internal(i, 1), :),'MarkerSize',20);

         end

     end

     hold off;
     pause;
end