SAR_Change_Detection_FDA_RMG/main.m at master · summitgao/SAR_Change_Detection_FDA_RMG · 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

%======================================================
% The files are the MATLAB source code for the paper:
%
% Feng Gao, Xiao Wang, Junyu Dong, Shengke Wang.
%
% Synthetic aperture radar image change detection based on
% frequencydomain analysis and random multigraphs
%
% Journal of Applied Remote Sensing. 12(1), 016010 (2018).
%
% The demo has not been well organized.
% Please contact me if you meet any problems.
%
% Email: gaofeng@ouc.edu.cn
%=======================================================


clear; clc;

addpath('./msqf/');
addpath('./rmg/');


%========== basic parameters =============
SmoothVal = 0.01;
PatSize = 5;
t_sal = 0.1;
%=========================================

im1   = imread('./pic/san_1.bmp');
im2   = imread('./pic/san_2.bmp');
im_gt = imread('./pic/san_gt.bmp');

im1   = double(im1(:,:,1));
im2   = double(im2(:,:,1));
im_gt = double(im_gt(:,:,1));

% obtain the difference image
fprintf(' ... ... compute the difference image ... ...\n');
im_di = di_gen(im1, im2);
im_di = mat2gray(im_di)*255;
im_di(:,:,2)=im_di(:,:,1); im_di(:,:,3)=im_di(:,:,1);


% compute the saliency map of the difference image
sal_map = sal_compute(SmoothVal, im_di);
sal_map = mat2gray(sal_map);
sal_map = double(sal_map > t_sal);

% find salient region
sal_idx = find(sal_map==1);
feat_vec = im_di(sal_idx);
fprintf(' ... ... clustering for sample selection begin ... ...\n');
feat_lab = gao_clustering(feat_vec);
fprintf(' ... ... clustering for sample selection finished !!!!\n');

clear i feat_vec SmoothVal;

% 获取 lab 信息
pos_idx = find(feat_lab == 1);
neg_idx = find(feat_lab == 0);
mid_idx = find(feat_lab == 0.5);

[ylen, xlen] = size(im1);

% 图像周围填零，然后每个像素周围取Patch，保存
mag = (PatSize-1)/2;
imTmp = zeros(ylen+PatSize-1, xlen+PatSize-1);
imTmp((mag+1):end-mag,(mag+1):end-mag) = im1;
im1 = im2col_general(imTmp, [PatSize, PatSize]);
imTmp((mag+1):end-mag,(mag+1):end-mag) = im2;
im2 = im2col_general(imTmp, [PatSize, PatSize]);
clear imTmp mag;
% 合并样本到 im
im1 = mat2imgcell(im1, PatSize, PatSize, 'gray');
im2 = mat2imgcell(im2, PatSize, PatSize, 'gray');
parfor idx = 1 : numel(im1)
    im_tmp = [im1{idx}; im2{idx}];
    im(idx, :) = im_tmp(:);
end
clear im1 im2 idx;

% 取出显著性区域的特征
im = im(sal_idx, :);


Dim = PatSize*PatSize*2;
kg = 2;
% 使用 random multi-graph 训练与测试
fprintf(' ... ... Graph number:%d ... ...\n', kg);
kf = floor(Dim/4);
% 得到训练样本在整体样本集中的 index
label_index = [pos_idx; neg_idx];
% 整个样本集的 groud truth index
labels = reshape(im_gt(sal_idx), length(sal_idx), 1) / 255 + 1;
labels(mid_idx) = 1;
[G,F]  = MultiGraphs(im,labels,label_index,kg,kf);


% obtain the classification results
[val, predict_res]=max(F,[],2);
predict_res = predict_res-1;

res = zeros(ylen, xlen);
for i = 1:length(sal_idx)
    res(sal_idx(i)) = predict_res(i);
end

[FP,FN,OE,PCC] = GetAccuracy(im_gt, res);
% save the results
fid = fopen('rec.txt', 'a');
fprintf(fid, 'ta = %f\n', t_sal);
fprintf(fid, 'PatSize = %d\n', PatSize);
fprintf(fid, 'FP: %d \n', FP);
fprintf(fid, 'FN: %d \n', FN);
fprintf(fid, 'OE: %d \n', OE);
fprintf(fid, 'PCC:   %f \n\n\n', PCC);
fclose(fid);

fprintf(' ===== Written change detection results to Res.txt ====\n\n');