forked from HuimingJia/face_recognition_matlab
-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathrecognization.m
executable file
·139 lines (116 loc) · 3.7 KB
/
recognization.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
function [] = recognization( inputFace)
%UNTITLED3 此处显示有关此函数的摘要
% 此处显示详细说明
global face_show;
global face_analysis;
global face_r_show;
global face_r_analysis;
imshow(inputFace,'parent',face_show);
%读入所有人脸
%每个人脸图像一列，存入faceLibrary_mat和faceInput_mat
num=112; %共有107个人
GNum = 1; %每个人有5个注册人脸
PNum = 1; %每个人有1个测试人脸
faceLibrary_mat = zeros(10000, num*GNum); %空矩阵, 注册集
faceInput_mat = zeros(10000, num*PNum); %空矩阵，测试集
%读入所有人脸数据
for person_id = 1:num
%逐一读注册集中的样本
%for gallery_id=1:GNum %每个人注册集中有5个人脸图像
filename = ['facelibrary/', num2str(person_id) ,'.png'];
im = imread(filename);
% if(person_id==107)
% disp(size(im,1));
% disp(size(im,2));
% end
im = double( reshape(im, 10000, 1) ); %将读入的100x100矩阵转为 10000x1的列向量
faceLibrary_mat(:,person_id) = im; %将之放入第person_id列
%end
end
%读测试集中的样本
im = double( reshape(inputFace, 10000, 1) ); %将读入的100x100矩阵转为 10000x1的列向量
faceInput_mat(:,1) = im;
%disp(['Person ID ', num2str(person_id)]);
disp('all samples have been loaded.');
%计算 distance matrix，矩阵元素是两个样本的距离
%距离越小说明越相似
% 提取LBP特征，然后计算Chi-square距离
disp('方法三：LBP');
faceLibrary_mat_lbp=[];
faceInput_mat_lbp=[];
%为所有注册集中的样本提取LBP特征
for idx = 1:size(faceLibrary_mat,2)
face_image = reshape(faceLibrary_mat(:,idx), 100,100);
faceLibrary_mat_lbp = [faceLibrary_mat_lbp, lbp7x7hist( face_image )];
end
face_image = reshape(faceInput_mat(:,1), 100,100);
faceInput_mat_lbp = [faceInput_mat_lbp, lbp7x7hist( face_image )];
weights=[ones(59,1)*2; %line 1
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*2;
ones(59,1)*2; %line 2
ones(59,1)*4;
ones(59,1)*4;
ones(59,1)*1;
ones(59,1)*4;
ones(59,1)*4;
ones(59,1)*2;
ones(59,1)*1; %line 3
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0; %line 4
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0;
ones(59,1)*0; %line 5
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0;
ones(59,1)*0; %line 6
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*2;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0;
ones(59,1)*0; %line 7
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*1;
ones(59,1)*0 ];
%下面代码将权重一次性乘到所有特征向量上，可以加快计算速度
faceLibrary_mat_wlbp=faceLibrary_mat_lbp .* repmat(weights,1, size(faceLibrary_mat_lbp,2));
faceInput_mat_wlbp=faceInput_mat_lbp .* repmat(weights,1, size(faceInput_mat_lbp,2));
%计算注册集中所有样本，与测试集中所有样本两两之间的距离，!!!卡方距离!!!
dist_mat = cal_dist_mat( faceLibrary_mat_lbp , faceInput_mat_lbp, 3 );
disp(dist_mat(:,1));
min= dist_mat(1,1);
count=1;
for i=1:size(dist_mat,1)
if(dist_mat(i,1)<=min)
count=i;
disp(count);
end
end
result_filename = ['facelibrary/',num2str(count) ,'.png'];
result_face_image = imread(result_filename);
imshow(result_face_image,'parent',face_r_show);
bar(faceInput_mat_lbp,'parent',face_analysis);
bar(faceLibrary_mat_lbp(:,count),'parent',face_r_analysis);
end