Fork me on GitHub

OpenCV之Practice

发表于 | 分类于
字数统计 4.5k 字 | 阅读时长 24 分钟

  1. 计算两个矩阵的相似性

  2. 画三角形、圆形检测出来

  3. Haar特征与Viola-Joines人脸检测

  4. 基于OpenCV的运动物体追踪

  5. 基于Opencv的数字图像识别

计算两个矩阵的相似性

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

//计算两个图的相似性
#include <iostream>
#include <opencv2/opencv.hpp>
using namespace std;
using namespace cv;

Mat srcImage, dstImage;

// 函数说明
// 这个函数是用来返回两个黑白图片的相似度的
// 输入是两个图片，输出是相似度
float isSimilar(Mat &src, Mat &dst)
{
	// 先定义两个图片的矩阵
	Mat src_copy, dst_copy;
	// resize(dst, src.row, src.col);
	// 讲两个图片缩放到src的尺寸上，并且用心的矩阵来储存
	resize(src, src_copy, src.size(), 0, 0, INTER_LINEAR);
	resize(dst, dst_copy, src.size(), 0, 0, INTER_LINEAR);


	float same = 0, diff = 0;
	// 两层for循环访问所有的元素点
	for (int i = 0; i < src_copy.rows; i++)
	{
		for (int j = 0; j < src_copy.cols; j++)
		{
			// 如果像素相同，那么same加1
			// 这里的代码明明是学习笔记上的，要好好看啊
			if (src_copy.at<uchar>(i, j) == dst_copy.at<uchar>(i, j))
				same += 1;
			// 如果像素不同，那么diff加1
			else
				diff += 1;
		}
	}
	// 返回相似度
	return same / (same + diff);
}

int main()
{
	// 在main函数里面读入两个图片
	srcImage = imread("8.png");
	dstImage = imread("9.png");
	// 对两个图片转换成灰度
	cvtColor(srcImage, srcImage, CV_BGR2GRAY);
	cvtColor(dstImage, dstImage, CV_BGR2GRAY);
	// 将两个图片转成黑白图，阈值化
	threshold(srcImage, srcImage, 100, 255, CV_THRESH_BINARY);
	threshold(dstImage, dstImage, 100, 255, CV_THRESH_BINARY);
	// 定义一个float
	float sim = 0;
	sim = isSimilar(srcImage, dstImage);

	cout << " the sim is " << sim << endl;
	system("pause");
	waitKey(0);
	return 0;
}

画三角形、圆形检测出来

任务说明：
首先创建一个512×512的3通道的白色的图像
然后在这个图像上画三角形，四角形，五角形，六角形，圆，以及椭圆
对这个图像，使用霍夫曼检查直线和圆

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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
#include <iostream>
#include <opencv2/opencv.hpp>

using namespace std;
using namespace cv;


//绘制椭圆
void drawEllipse(Mat img, double angle)
{
	int thickness = 2;
	int lineType = 8;

	ellipse(img, Point(40, 40), Size(20, 40), angle, 0, 360, Scalar(255, 129, 0), thickness, lineType);
}

//绘制圆
void drawCircle(Mat img, Point center)
{
	int thickness = 2;
	int lineType = 8;

	circle(img, center, 30, Scalar(168, 109, 255), thickness, lineType);
}

//画线
void drawLine(Mat img, Point start, Point end)
{
	int thickness = 2;
	int lineType = 8;

	line(img, start, end, Scalar(0, 0, 0), thickness, lineType);
}

//绘制矩形
void drawRectangle(Mat img, Point pt1, Point pt2)
{
	int thickness = 2;
	int lineType = 8;

	rectangle(img, pt1, pt2, Scalar(0, 255, 0), thickness, lineType);
}

//绘制三角形
void drawLine_Triangle(Mat img)
{
	int thickness = 2;
	int lineType = 8;

	Point l1_start(200, 150), l1_end(50, 250);
	Point l2_start(50, 250), l2_end(400, 380);
	Point l3_start(400, 380), l3_end(200, 150);
	line(img, l1_start, l1_end, Scalar(0, 0, 0), thickness, lineType);
	line(img, l2_start, l2_end, Scalar(0, 0, 0), thickness, lineType);
	line(img, l3_start, l3_end, Scalar(0, 0, 0), thickness, lineType);
}

////绘制多边形
//void drawfillPloy(Mat img)
//{
//	int lineType = 8;
//
//	Point rookPoints[1][3];
//	rookPoints[0][0] = Point(30, 30);
//	rookPoints[0][1] = Point(60, 60);
//	rookPoints[0][2] = Point(90, 90);
//
//	Point *pts[1] = {rookPoints[0]}; //点序列的序列
//	int npts[] = {3};   //pts[i]中点的数目
//	int ncontours = 3;  //pts中的序列数
//
//	fillPoly(img, pts, npts, ncontours, Scalar(168, 200, 120), lineType);
//}

//使用霍夫线检测出线
//Mat check_line(Mat img)
//{
//	Mat dstImage;
//
//	cvtColor(img, dstImage, COLOR_BGR2GRAY);
//	adaptiveThreshold(dstImage, dstImage, 50, ADAPTIVE_THRESH_GAUSSIAN_C, THRESH_BINARY_INV, 10, 10);
//
//	vector<Vec2f> lines;
//	HoughLines(dstImage, lines, 1, CV_PI/180, 80, 0, 0);
//
//	vector<Vec2f> ::iterator itl = lines.begin();
//	while (itl != lines.end())
//	{
//		rectangle(img, Point((*itl)[0] + 5, (*itl)[1] + 5), Point((*itl)[2] + 5, (*itl)[3] + 5), 2, 8);
//		++itl;
//	}
//
//	imshow("check_lines", img);
//
//	return img;
//}

//使用霍夫线检测出线
Mat check_line(Mat& img)
{
	Mat dstImage;

	cvtColor(img, dstImage, COLOR_BGR2GRAY);

	Canny(dstImage, dstImage, 150, 255, 3);

	//使用霍夫曼检测
	//vector<Vec4f> lines;
	// 注意这里的第二个参数
	//HoughLines(dstImage, lines, 1, CV_PI/180.0, 80,50,10);

	// 以下修改为霍夫曼概率直线检测函数
	// 注意这里的参数
	vector<Vec4i> lines;
	// 注意一下，使用霍夫曼概率直线检测与霍夫曼直线检测的区别
	// 注意这里的第二个参数
	HoughLinesP(dstImage, lines, 1, CV_PI / 180.0, 80, 50, 10);

	Scalar color = Scalar(0, 0, 255);
	for (size_t i = 0; i < lines.size(); i++) {
		cout << " i is " << i << "  ";
		Vec4i hline = lines[i];
		cout << hline[0] << " ||| " << hline[1] << " ||| " << hline[2] << " |||| " << hline[3] << endl;
		line(img, Point(hline[0], hline[1]), Point(hline[2], hline[3]), color, 3, LINE_AA);
	}

	imshow("check_lines", img);

	return img;
}

//使用霍夫圆检测出圆形
Mat check_circle(Mat img)
{
	Mat dstImage;

	cvtColor(img, dstImage, COLOR_BGR2GRAY);
	GaussianBlur(dstImage, dstImage, Size(7, 7), 1.5);

	vector<Vec3f> circles;
	HoughCircles(dstImage, circles, CV_HOUGH_GRADIENT, 2, 30, 200, 100, 20, 80);

	/*vector<Vec3f> ::iterator itc = circles.begin();
	while (itc != circles.end())
	{
		circle(img, Point((*itc)[0], (*itc)[1]), (*itc)[2] + 20, Scalar(0, 0, 255), 6);
		++itc;
	}*/

	for (size_t i = 0; i < circles.size(); i++)
	{
		Point center(cvRound(circles[i][0]), cvRound(circles[i][1]));
		int r = cvRound(circles[i][2]);

		circle(img, center, r, Scalar(0, 0, 255), 3, 8, 0);
	}

	imshow("check_Circle", img);

	return img;
}

int main()
{
	Mat srcImage(512, 512, CV_8UC3 ,Scalar(255, 255, 255));

	//imshow("srcImage", srcImage);

	//绘制椭圆形
	drawEllipse(srcImage, 30);
	//绘制圆形
	Point center(300, 100);
	drawCircle(srcImage, center);
	//绘制线段
	Point start(100, 100);
	Point end(200, 200);
	drawLine(srcImage, start, end);
	//绘制矩形
	Point pt1(300, 300); //矩形的第一个顶点
	Point pt2(500, 500); //矩形的对角线顶点
	drawRectangle(srcImage, pt1, pt2);
	//绘制三角形
	drawLine_Triangle(srcImage);

	//putText(srcImage, "dasha is wangjiacun", Point(70, 70), FONT_HERSHEY_PLAIN, 2, 255, 2);

	imshow("out", srcImage);
	                                                                   
	check_line(srcImage);
	check_circle(srcImage);

	waitKey(0);
	return 0;
}

Haar特征与Viola-Joines人脸检测

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
//OpenCV3/C++ Haar特征与Viola-Joines人脸检测
#include <iostream>
#include <opencv2/opencv.hpp>
#include <opencv2/objdetect.hpp>
using namespace std;
using namespace cv;

int main()
{
	Mat srcImage;
	CascadeClassifier face_cacade;

	VideoCapture capture(0);
	if (!face_cacade.load("D:\\Project_OpenCV\\Cmake\\install\\etc\\haarcascades\\haarcascade_frontalface_alt.xml"))
	{
		printf("can not load the file...\n");
		return -1;
	}

	while (1)
	{
		capture >> srcImage;
		vector<Rect> faces;
		Mat gray_image;
		cvtColor(srcImage, gray_image, CV_BGR2GRAY);
		//直方图均衡化
		equalizeHist(gray_image, gray_image);
		face_cacade.detectMultiScale(gray_image, faces, 1.1, 2, 0 | CV_HAAR_SCALE_IMAGE, Size(1, 1));
		//框选出脸部区域
		for (int i = 0; i < faces.size(); i++)
		{
			RNG rng(i);
			Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), 20);
			rectangle(srcImage, faces[static_cast<int>(i)], color, 2, 8, 0);
		}
		imshow("face", srcImage);
		waitKey(30);	//延时30
	}



	Mat srcImage;

	srcImage = imread("7.jpg");

	CascadeClassifier face_cacade;

	VideoCapture capture(0);
	if (!face_cacade.load("D:\\Project_OpenCV\\Cmake\\install\\etc\\haarcascades\\haarcascade_frontalface_alt.xml"))
	{
		printf("can not load the file...\n");
		return -1;
	}
	vector<Rect> faces;
	Mat gray_image;
	cvtColor(srcImage, gray_image, CV_BGR2GRAY);

	//直方图均衡化
	equalizeHist(gray_image, gray_image);
	face_cacade.detectMultiScale(gray_image, faces, 1.1, 2, 0 | CV_HAAR_SCALE_IMAGE, Size(1, 1));

	//框选出脸部区域
	for (int i = 0; i < faces.size(); i++)
	{
		RNG rng(i);
		Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), 20);
		rectangle(srcImage, faces[static_cast<int>(i)], color, 2, 8, 0);
	}

	imshow("face", srcImage);
	waitKey(0);


	return 0;
}

基于OpenCV的运动物体追踪

给一段视频，程序对视频中运动的物体进行追踪，并用框框起来

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
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
////1.运动物体检测----背景减法
//#include <iostream>
//#include <opencv2/opencv.hpp>
//using namespace std;
//using namespace cv;
//
////运动物体检测函数声明
//Mat MoveDetect(Mat backgroung, Mat frame);
//
//int main()
//{
//
//	VideoCapture video("move.avi");
//	if (!video.isOpened())
//	{
//		printf("Oh, no, video is error");
//		return -1;
//	}
//
//	int frameCount = video.get(CV_CAP_PROP_FRAME_COUNT); //获取帧数
//	double FPS = video.get(CV_CAP_PROP_FPS);//获取FPS
//	Mat frame; //存储帧
//	Mat background; //存储背景图像
//	Mat result; //存储结果图像
//
//	for (int i = 0; i < frameCount; i++)
//	{
//		video >> frame; //读帧进frame
//		imshow("frame", frame);
//		if (frame.empty()) //对帧进行异常检测
//		{
//			cout << "frame is empty" << endl;
//			break;
//		}
//		int framePosition = video.get(CV_CAP_PROP_POS_FRAMES); //获取帧位置(第几帧)
//		cout << "framePosition: " << framePosition << endl;
//		if (framePosition == 1) //将第一帧作为背景图像
//			background = frame.clone();
//		result = MoveDetect(background, frame); //调用MoveDect()进行运动物体检测，返回值存入result
//		imshow("result", result);
//		if (waitKey(1000.0 / FPS) == 27) //按原FPS显示
//		{
//			cout << "ESC退出！" << endl;
//			break;
//		}
//	}
//
//
//	return 0;
//}
//
//Mat MoveDetect(Mat background, Mat frame)
//{
//	Mat result = frame.clone();
//	//1.将background和frame转为灰度图
//	Mat gray1, gray2;
//	cvtColor(background, gray1, CV_BGR2GRAY);
//	cvtColor(frame, gray2, CV_BGR2GRAY);
//	//2.将background和frame做差
//	Mat diff;
//	absdiff(gray1, gray2, diff);
//	imshow("diff", diff);
//	//3. 对差值图diff_thresh进行阈值化处理
//	Mat diff_thresh;
//	threshold(diff, diff_thresh, 50, 255, CV_THRESH_BINARY);
//	imshow("diff_thresh", diff_thresh);
//	//4.腐蚀
//	Mat kernel_erode = getStructuringElement(MORPH_RECT, Size(3, 3));
//	Mat kernel_dilate = getStructuringElement(MORPH_RECT, Size(15, 15));
//	erode(diff_thresh, diff_thresh, kernel_erode);
//	imshow("erode", diff_thresh);
//	//5.膨胀
//	dilate(diff_thresh, diff_thresh, kernel_dilate);
//	imshow("dilate", diff_thresh);
//	//6.查找轮廓并绘制轮廓
//	vector<vector<Point>> contours;
//	findContours(diff_thresh, contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
//	drawContours(result, contours, -1, Scalar(0, 0, 255), 2);//在result上绘制轮廓
//	//7.查找正外接矩形
//	vector<Rect> boundRect(contours.size());
//	for (int i = 0; i < contours.size(); i++)
//	{
//		boundRect[i] = boundingRect(contours[i]);
//		rectangle(result, boundRect[i], Scalar(0, 255, 0), 2); //在result上绘制正外接矩阵
//	}
//
//	return result; //返回result
//}

#include "opencv2/opencv.hpp"
#include<iostream>
using namespace std;
using namespace cv;

Mat MoveDetect(Mat frame1, Mat frame2)
{
	Mat result = frame2.clone();
	Mat gray1, gray2;
	cvtColor(frame1, gray1, CV_BGR2GRAY);
	cvtColor(frame2, gray2, CV_BGR2GRAY);

	Mat diff;
	absdiff(gray1, gray2, diff);
	imshow("absdiss", diff);
	threshold(diff, diff, 45, 255, CV_THRESH_BINARY);
	imshow("threshold", diff);

	Mat element = getStructuringElement(MORPH_RECT, Size(3, 3));
	Mat element2 = getStructuringElement(MORPH_RECT, Size(25, 25));
	erode(diff, diff, element);
	imshow("erode", diff);

	dilate(diff, diff, element2);
	imshow("dilate", diff);

	vector<vector<Point>> contours;
	vector<Vec4i> hierarcy;
	//画椭圆及中心
	findContours(diff, contours, hierarcy, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
	cout << "num=" << contours.size() << endl;
	vector<RotatedRect> box(contours.size());
	for (int i = 0; i<contours.size(); i++)
	{
		box[i] = fitEllipse(Mat(contours[i]));
		ellipse(result, box[i], Scalar(0, 255, 0), 2, 8);
		circle(result, box[i].center, 3, Scalar(0, 0, 255), -1, 8);
	}
	return result;
}

void main()
{
	VideoCapture cap("move.avi");
	if (!cap.isOpened()) //检查打开是否成功
		return;
	Mat frame;
	Mat result;
	Mat background;
	int count = 0;
	while (1)
	{
		cap >> frame;
		if (frame.empty())
			break;
		else {
			count++;
			if (count == 1)
				background = frame.clone(); //提取第一帧为背景帧
			imshow("video", frame);
			result = MoveDetect(background, frame);
			imshow("result", result);
			if (waitKey(50) == 27)
				break;
		}
	}
	cap.release();
}


////2.运动物体检测----帧查法
//#include <iostream>
//#include <opencv2/opencv.hpp>
//using namespace std;
//using namespace cv;
//
//Mat MoveDetect(Mat background, Mat frame);
//
//int main()
//{
//	VideoCapture video("move.avi");
//	if (!video.isOpened())
//	{
//		printf("Oh, no, video is error");
//		return -1;
//	}
//
//	int frameCount = video.get(CV_CAP_PROP_FRAME_COUNT);
//	double FPS = video.get(CV_CAP_PROP_FPS);
//	Mat frame;
//	Mat background;
//	Mat result;
//
//	for(int i = 0; i < frameCount; i++)
//	{
//		video >> frame;
//		imshow("frame", frame);
//		if (frame.empty())
//		{
//			printf("frame is empty!");
//			break;
//		}
//
//		if (i == 0)
//		{
//			result = MoveDetect(frame, frame);
//		}
//		else
//		{
//			result = MoveDetect(background, frame);
//		}
//
//		imshow("result", result);
//		if (waitKey(1000.0 / FPS) == 27)
//			break;
//		background = frame.clone();
//	}
//
//	return 0;
//}
//
//Mat MoveDetect(Mat background, Mat frame)
//{
//	Mat result = frame.clone();
//	Mat gray1, gray2;
//	cvtColor(background, gray1, CV_BGR2GRAY);
//	cvtColor(frame, gray2, CV_BGR2GRAY);
//
//	Mat diff;
//	absdiff(gray1, gray2, diff);
//	imshow("diff", diff);
//
//	Mat diff_thresh;
//	threshold(diff, diff_thresh, 50, 255, CV_THRESH_BINARY);
//	imshow("threshold", diff_thresh);
//
//	Mat kernel_erode = getStructuringElement(MORPH_RECT, Size(3, 3));
//	Mat kernel_dilate = getStructuringElement(MORPH_RECT, Size(18, 18));
//
//	erode(diff_thresh, diff_thresh, kernel_erode);
//	imshow("erode", diff_thresh);
//	dilate(diff_thresh, diff_thresh, kernel_dilate);
//	imshow("dilate", diff_thresh);
//
//	vector<vector<Point>> contours;
//	findContours(diff_thresh, contours, CV_RETR_EXTERNAL, CV_CHAIN_APPROX_NONE);
//	drawContours(result, contours, -1, Scalar(0, 0, 255), 2);
// 
//	vector<Rect> boundRect(contours.size());
//	for (int i = 0; i < contours.size(); i++)
//	{
//		boundRect[i] = boundingRect(contours[i]);
//		rectangle(result, boundRect[i], Scalar(0, 255, 0), 2);
//	}
//	return result;
//}

基于Opencv的数字图像识别

准备：
1.10个模板图像
2.相似度检测函数、轮廓检测函数
具体内容：
1.读入10个模板图像
2.对目标图像进行轮廓检测并剪切出识别的数字
3.对剪切出的数字与10个模板图像相似度计算挑选出最相似的

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
#include <iostream>
#include <opencv2/opencv.hpp>
#include <vector>
#include <algorithm>
using namespace std;
using namespace cv;

Mat srcImage, dstImage;
vector<Mat> vec;
vector<Mat> tutu;

//相似性函数
float isSimilar(Mat &src, Mat &dst)
{
	// 先定义两个图片的矩阵
	Mat src_copy, dst_copy;
	// resize(dst, src.row, src.col);
	// 讲两个图片缩放到src的尺寸上，并且用新的矩阵来储存
	resize(src, src_copy, src.size(), 0, 0, INTER_LINEAR);
	resize(dst, dst_copy, src.size(), 0, 0, INTER_LINEAR);


	float same = 0, diff = 0;
	// 两层for循环访问所有的元素点
	for (int i = 0; i < src_copy.rows; i++)
	{
		for (int j = 0; j < src_copy.cols; j++)
		{
			// 如果像素相同，那么same加1
			// 这里的代码明明是学习笔记上的，要好好看啊
			if (src_copy.at<uchar>(i, j) == dst_copy.at<uchar>(i, j))
				same += 1;
			// 如果像素不同，那么diff加1
			else
				diff += 1;
		}
	}
	// 返回相似度
	return same / (same + diff);
}
//轮廓检测函数
void isContours(Mat &src)
{
	Mat img_gray;
	resize(src, img_gray, src.size(), 0, 0, INTER_LINEAR);

	cvtColor(src, img_gray, CV_BGR2GRAY);

	// 注意在取阈值的时候，最后一个参数
	threshold(img_gray, dstImage, 100, 255, CV_THRESH_BINARY);

	vector<vector<Point> > contours;
	vector<Vec4i> hierarchy;
	// 注意这里的第四个参数，决定了检测的准确性
	findContours(dstImage, contours, hierarchy, RETR_EXTERNAL, CHAIN_APPROX_SIMPLE, Point(0, 0));

	RNG rng(0);
	for (int i = 0; i < contours.size(); i++)
	{
		Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
		Rect rect = boundingRect(contours[i]);//检测外轮廓
		Mat temp = src(rect);

		imwrite("shacun" + to_string(i) + ".png", temp);
		rectangle(src, rect, Scalar(0, 0, 255), 3);//对外轮廓加矩形框
	}
}
int main()
{
	//调用轮廓检测函数
	srcImage = imread("digits_4.jpg");
	if (!srcImage.data)
	{
		printf("Oh, no, srcImage is error");
		return 0;
	}
	// 轮廓检测并保存数据
	isContours(srcImage);


	//调用相似性函数
	for (int i = 0; i < 10; i++)
	{
		//模板10个图像
		Mat s_src = imread("tutu" + to_string(i) + ".png");
		vec.push_back(s_src);

		//切割的10个图像
		Mat tu = imread("shacun" + to_string(i) + ".png");
		tutu.push_back(tu);
	}
	for (int j = 0; j < 10; j++)
	{
		int val = -1;
		float maxSim = 0;
		for (int i = 0; i < 10; i++)
		{
			float sim = 0;
			sim = isSimilar(vec[i], tutu[j]);
			if (maxSim <= sim)
			{
				maxSim = sim;
				val = i;
			}
		}
		imshow("aim is ", tutu[j]);
		imshow("model is ", vec[val]);
		waitKey(0);
		cout << " the val is " << val << " the sim is  " << maxSim << endl;
	}
	

	system("pause");
	return 0;
}
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
147
148
#include <iostream>
#include <opencv2/opencv.hpp>
#include <vector>
#include <algorithm>
using namespace std;
using namespace cv;

Mat srcImage, dstImage;
vector<Mat> vec;


/**
代码如何理解
从main函数开始，首先调用init函数
我们用init函数来加载模板图像到全局的数组vec里面
然后调用isContours函数，使用isContours函数对目标图片检测
检测完成后，接着就调用predict函数来预测轮廓检测的数数字
在predict函数里面，我们调用isSimilar函数来检测出最相似的模板图片
然后，输出对应的数字
*/


// 函数声明
void init();
void isContours(Mat &src);
float isSimilar(Mat &src, Mat &dst);
void predict(Mat &dst);

// 初始化加载模板函数
void init()
{
	// 初始化函数，用来加载模板图片
	for (int i = 0; i < 10; i++)
	{
		//模板10个图像
		Mat s_src = imread("tutu" + to_string(i) + ".png");
		vec.push_back(s_src);
	}
	cout << "***************" << "init is OK." << "***************" << endl;

}

// 预测函数
void predict(Mat &dst)
{
	// 检测数字函数
	int val = -1;
	float maxSim = 0;
	for (int i = 0; i < 10; i++)
	{
		// cvtColor(vec[i], vec[i], CV_BGR2GRAY);
		// threshold(vec[j], vec[j], 100, 255, CV_THRESH_BINARY);

		float sim = 0;
		sim = isSimilar(vec[i], dst);
		if (maxSim <= sim)
		{
			maxSim = sim;
			val = i;
		}
	}
	imshow("aim is ", dst);
	imshow("model is ", vec[val]);
	waitKey(0);
	cout << " the val is " << val << " the sim is  " << maxSim << endl;

}


//轮廓检测函数
void isContours(Mat &src)
{
	Mat img_gray;
	resize(src, img_gray, src.size(), 0, 0, INTER_LINEAR);

	cvtColor(src, img_gray, CV_BGR2GRAY);

	// 注意在取阈值的时候，最后一个参数
	threshold(img_gray, dstImage, 100, 255, CV_THRESH_BINARY);

	// imshow("thre",dstImage);

	vector<vector<Point> > contours;
	vector<Vec4i> hierarchy;
	// 注意这里的第四个参数，决定了检测的准确性
	findContours(dstImage, contours, hierarchy, RETR_EXTERNAL, CHAIN_APPROX_SIMPLE, Point(0, 0));

	RNG rng(0);
	for (int i = 0; i < contours.size(); i++)
	{
		Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
		Rect rect = boundingRect(contours[i]);//检测外轮廓
	    // rectangle(src, rect, Scalar(0, 0, 255), 3);//对外轮廓加矩形框
		Mat dst = src(rect);
		// 预测函数
		predict(dst);
		// imwrite("../shacun/" + to_string(i) + ".png", temp);
	}
	// imshow("src",src);
	// waitKey(0);
}


//相似性函数
float isSimilar(Mat &src, Mat &dst)
{
	// 先定义两个图片的矩阵
	Mat src_copy, dst_copy;
	// resize(dst, src.row, src.col);
	// 讲两个图片缩放到src的尺寸上，并且用新的矩阵来储存
	resize(src, src_copy, src.size(), 0, 0, INTER_LINEAR);
	resize(dst, dst_copy, src.size(), 0, 0, INTER_LINEAR);

	float same = 0, diff = 0;
	// 两层for循环访问所有的元素点
	for (int i = 0; i < src_copy.rows; i++)
	{
		for (int j = 0; j < src_copy.cols; j++)
		{
			// 如果像素相同，那么same加1
			// 这里的代码明明是学习笔记上的，要好好看啊
			if (src_copy.at<uchar>(i, j) == dst_copy.at<uchar>(i, j))
				same += 1;
			// 如果像素不同，那么diff加1
			else
				diff += 1;
		}
	}
	// 返回相似度
	return same / (same + diff);
}


int main()
{
	// init初始化函数加载模板图片
	init();
	//调用轮廓检测函数
	srcImage = imread("digits_4.jpg");
	if (!srcImage.data)
	{
		printf("Oh, no, srcImage is error");
		return 0;
	}
	// 轮廓检测并保存数据
	isContours(srcImage);
	return 0;
}
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
////基于OpenCV的数字识别
//#include <iostream>
//#include <opencv2/opencv.hpp>
//#include <vector>
//using namespace std;
//using namespace cv;
//
//Mat srcImage, dstImage;
//Mat imgROI;
//
//int main()
//{
//	srcImage = imread("digits_4.jpg");
//	if (!srcImage.data)
//	{
//		printf("Oh, no, srcImage is error");
//		return false;
//	}
//
//	imshow("srcImage", srcImage);
//
//	cvtColor(srcImage, srcImage, CV_BGR2GRAY);
//
//	threshold(srcImage, dstImage, 100, 255, CV_THRESH_BINARY);
//
//	/*Mat kernel_erode = getStructuringElement(MORPH_RECT, Size(3, 3));
//	Mat kernel_dilate = getStructuringElement(MORPH_RECT, Size(15, 15));
//
//	erode(dstImage, dstImage, kernel_erode);
//	dilate(dstImage, dstImage, kernel_dilate);*/
//
//
//	vector<vector<Point> > contours;
//	vector<Vec4i> hierarchy;
//	findContours(dstImage, contours, RETR_TREE, CHAIN_APPROX_SIMPLE, Point(0, 0));
//
//	RNG rng(0);
//	for (int i = 0; i < contours.size(); i++)
//	{
//		Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
//		drawContours(dstImage, contours, i, color, 2, 8, hierarchy, 0, Point(0, 0));
//		Rect rect = boundingRect(contours[i]);
//
//		//imgROI = dstImage(Rect(rect.x, rect.y, rect.width, rect.height ));
//
//		//imwrite("imgROI" + to_string(i) + ".png", imgROI);
//		cout << " the area is " << contourArea(contours[i]) << endl;
//		cout << "contours" << i << " height =" << rect.height << " width = " << rect.width << " rate = " << ((float)rect.width / rect.height) << endl;
//		Mat temp = srcImage(rect);
//		// imshow("tmp ",temp);
//		// cout << " temp shape is " << rect.x << " " << rect.y << " " << endl;
//
//		if(rect.height >= 60 && rect.width >= 30)
//		{
//			imwrite("digits" + to_string(i) + ".png", temp);
//		}
//		
//		rectangle(dstImage, rect, Scalar(0, 0, 255), 3);//对外轮廓加矩形框
//
//	}
//
//	imshow("dstImage", dstImage);
//	
//
//	waitKey(0);
//
//	return 0;
//}


#include <iostream>
#include <opencv2/opencv.hpp>
#include <vector>
using namespace std;
using namespace cv;

Mat srcImage, dstImage;
Mat dst;

int main()
{
	srcImage = imread("digits_3.png");
	Mat img_gray;
	cvtColor(srcImage, img_gray, CV_BGR2GRAY);

	if (!srcImage.data)
	{
		printf("Oh, no, srcImage is error");
		return 0;
	}

	namedWindow("gray", CV_WINDOW_AUTOSIZE);
	imshow("gray", img_gray);

	// 注意在取阈值的时候，最后一个参数
	threshold(img_gray, dstImage, 100, 255, CV_THRESH_BINARY);
	//threshold(img_gray, dstImage, 100, 255, CV_THRESH_BINARY_INV);



	namedWindow("Threshold", CV_WINDOW_AUTOSIZE);
	imshow("Threshold", dstImage);

	vector<vector<Point> > contours;
	vector<Vec4i> hierarchy;
	// 注意这里的第四个参数，决定了检测的准确性
	findContours(dstImage, contours, hierarchy, RETR_EXTERNAL, CHAIN_APPROX_SIMPLE, Point(0, 0));

	// imshow("dstImage", dstImage);
	cout << srcImage.rows << " " << srcImage.cols;
	RNG rng(0);
	for (int i = 0; i < contours.size(); i++)
	{
		cout << " i is " << i << endl;
		Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
		// drawContours(srcImage, contours, i, color, 2, 8, hierarchy, 0, Point(0,0));
		Rect rect = boundingRect(contours[i]);//检测外轮廓
		cout << " the area is " << contourArea(contours[i]) << endl;
		cout << "contours" << i << "height=" << rect.height << "width =" << rect.width << "rate =" << ((float)rect.width / rect.height) << endl;
		Mat temp = srcImage(rect);
		// imshow("tmp ",temp);
		// cout << " temp shape is " << rect.x << " " << rect.y << " " << endl;
		// if(rect.height >= 60 && rect.width >= 30)
		imwrite("tutu" + to_string(i) + ".png", temp);
		rectangle(srcImage, rect, Scalar(0, 0, 255), 3);//对外轮廓加矩形框
	}

	namedWindow("output", CV_WINDOW_AUTOSIZE);
	imshow("output", srcImage);

	waitKey(0);

	return 0;
}