-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathMain.m
146 lines (132 loc) · 6.48 KB
/
Main.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
140
141
142
143
144
145
146
function result = Main(fullfilename,handles)
[B source]= Readimg(fullfilename,handles);%获取每个种子轮廓绝对坐标下的像素集合，B为所有种子边缘像素坐标集合，source为原始图像矩阵信息
result = source;%result为输出结果，根据需要随意指定
%load svmStruct
B_items_tangle = zeros(length(B),4);%该矩阵用于存储一张图中所有种子的矩形框坐标，1234位分别代表矩形框的上下左右像素位
img_with_rect = source;%该矩阵为带矩形框的图片矩阵
%img1 = figure(1);
axes(handles.axes_taged);
imshow(img_with_rect);%画图带矩形，
global svmStruct;%定义向量机全局变量
svmStruct = load('svmStruct.mat');%从mat文件中载入向量机训练结果
svmStruct = svmStruct.svmStruct;%为向量机变量赋值
for i=1:length(B)
B_max = max(B{i});%第i个种子边缘坐标最大值
B_min = min(B{i});%第i个种子边缘坐标最小值
B_items_tangle(i,1) = B_min(1);%1234位分别代表矩形框的上下左右像素位
B_items_tangle(i,2) = B_max(1);
B_items_tangle(i,3) = B_min(2);
B_items_tangle(i,4) = B_max(2);
for j=1:size(B{i},1)%将每个种子的边缘坐标由全局值转化为相对值，用于将每个种子单独以图片方式存储
B{i}(j,1) = B{i}(j,1) - B_min(1)+1;
B{i}(j,2) = B{i}(j,2) - B_min(2)+1;
end
B_max = max(B{i});%获得坐标转化后的种子坐标极值
B_min = min(B{i});
seed_rect_length = 70;%设置种子单独存储为图片时大小为70x70
if(B_max(1)+2>seed_rect_length || B_max(2)+2>seed_rect_length)%将种子图像移到70x70图片中央
if(B_max(1)>B_max(2))
seed_rect_length = B_max(1)+2;
else
seed_rect_length = B_max(2)+2;
end
end
B_item = zeros(seed_rect_length,seed_rect_length);%定义该矩阵用于存储种子二值图像信息
for j=1:size(B{i},1)
B_item(floor((seed_rect_length - B_max(1)+2)/2)+B{i}(j,1)+1,floor((seed_rect_length - B_max(2)+2)/2)+B{i}(j,2)+1) = 255;
end
assess = Judgeimg(rotateimg(B_item,B{i}));%该函数用向量机特征库判断当前种子好坏，assess为结果
axes(handles.axes_taged);
if(assess>0.9)%如果判断结果概率大于0.9，就框绿色
rectangle('Position',[B_items_tangle(i,3),B_items_tangle(i,1),B_items_tangle(i,4) - B_items_tangle(i,3)+1,B_items_tangle(i,2) - B_items_tangle(i,1)+1],'EdgeColor','g');
else%否则，就框红色
rectangle('Position',[B_items_tangle(i,3),B_items_tangle(i,1),B_items_tangle(i,4) - B_items_tangle(i,3)+1,B_items_tangle(i,2) - B_items_tangle(i,1)+1],'EdgeColor','r');
end%为种子编号
text(B_items_tangle(i,3),B_items_tangle(i,1),sprintf('%d',i),'color','g');
B_items{i,1} = B_item;%获取每个种子相对坐标下像素
% path = fullfile('C:\Users\Administrator\Desktop\corn-seeds\seeds',sprintf('%03d.jpg',296+i));
% imwrite(rotateimg(B_items{i},B{i}),path,'jpg');
end
%disp(B_items_tangle);
B_items_circum = zeros(length(B_items),1);%周长
B_items_area = zeros(length(B_items),1);%面积
B_items_area_per = zeros(length(B_items),1);%面积占本身矩形面积百分比
B_items_area_average_per = zeros(length(B_items),1);%面积占平均面积百分比
B_items_totol_area = 0;%各个种子的总面积
B_items_totol_tangle_area = 0;%各个种子所在矩形的总面积
items_num = length(B);%输入一下循环的种子数，目前取所有种子
for i = 1:items_num
%subplot(3,5,i),imshow(B_items{i});
B_items_circum(i,1) = length(find(B_items{i}==255));%计算种子周长放入B_items_circum中
B_item = B_items{i};
item_inside = B_item>150; % inside
item_inside = bwfill(item_inside,'holes');
item_inside = bwareaopen(item_inside,1000);
B_items_area(i,1) = length(find(item_inside==1));%计算种子面积放入B_items_area中
B_items_totol_area = B_items_totol_area+B_items_area(i,1);%累加当前图片中种子的总面积
B_items_area_per(i,1) = 100*B_items_area(i,1)./(size(B_item,1)*size(B_item,2));%计算当前种子面积占种子框区面积的百分比
B_items_totol_tangle_area = B_items_totol_tangle_area+size(B_item,1)*size(B_item,2);%累加计算所有种子框区的面积和
%subplot(3,5,i+5),imshow(item_inside),title('inside');
end
result = B_items_circum;%result为种子周长集合
% img2 = figure(2);
% figure(img2);
% plot(result);%输出种子周长曲线
for i = 1:items_num
B_items_area_average_per(i,1) = B_items_area(i,1)/(B_items_totol_tangle_area/items_num);%计算每个种子面积占平均框区面积的百分比
end
%imshow(B{1});
%imshow(B{2});
% imshow(rotateimg(B_items{1},B{1}));
% figure;
% ExerciseOval(B{3});
% figure;
% ExerciseOval(B{4});
% figure;
% ExerciseOval(B{5});
end
function alpha = ExerciseOval(source_border)%该函数输出alpha为当前种子经过椭圆拟合后算出的长轴与水平线夹角，用于后续旋转种子到水平或竖直位置
%?拟合椭圆型曲线段
%?设出圆锥曲线方程?
B_max = max(source_border);%种子边缘坐标最大值
B_min = min(source_border);%种子边缘坐标最小值
B_max(1) = -B_max(1);
B_min(1) = -B_min(1);
for j=1:size(source_border,1)%将种子边缘左边进行绕x轴翻转，并扩展种子画布大小为70x70
source_border(j,1) = -source_border(j,1);
source_border(j,1) = source_border(j,1) - B_min(1)+1+floor(((70-(B_max(1)-B_min(1)))/2));
source_border(j,2) = source_border(j,2) - B_min(2)+1+floor(((70-(B_max(2)-B_min(2)))/2));
end
F=@(p,x)p(1)*x(:,1).^2+p(2)*x(:,1).*x(:,2)+p(3)*x(:,2).^2+p(4)*x(:,1)+p(5)*x(:,2)+p(6);%椭圆一般方程
Up=source_border;%待拟合点集合
p0=[1 1 1 1 1 1];%p0系数初值
p=nlinfit(Up,zeros(size(Up,1),1),F,p0);%拟合系数，最小二乘方法
A=p(1)/p(6);
B=p(2)/p(6);
C=p(3)/p(6);
D=p(4)/p(6);
E=p(5)/p(6);
% X_center = (B*E-2*C*D)/(4*A*C - B^2);%椭圆中心点坐标
% Y_center = (B*D-2*A*E)/(4*A*C - B^2);
%长轴倾角
alpha=0.5 * atan(B/(A-C));%找到长轴倾角，以便旋转
end
function B = rotateimg(source,border_source)%旋转种子
angel = round(ExerciseOval(border_source)*180/pi);%得到旋转角度
big_source = imresize(source,[200,200]);%将原矩阵放大
B = imrotate(big_source,angel,'bilinear','crop');%旋转放大后的矩阵angle度
B = imresize(B,size(source));%还原旋转后的矩阵为初始大小
B = im2bw(B,0.99);%二值化去除毛刺
H = fspecial('unsharp');
B = imfilter(B,H,'replicate');%拉普拉斯锐化
end
% figure;
% imshow(B);
function result = Judgeimg(img_mat)%计算种子的可信度
%% 训练完成后保存 svmStruct即可对新输入的对象进行分类了无需再执行上面训练阶段代码
img=img_mat;
im=imresize(img,[64,64]);
hogt =hogcalculator(im);
global svmStruct;%全局向量机结构
result = svmclassify(svmStruct,hogt);%result的值即为分类结果
end