用opencv中svm分割图像

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
#include <opencv2/core.hpp>
#include <opencv2/imgproc.hpp>
#include "opencv2/imgcodecs.hpp"
#include <opencv2/highgui.hpp>
#include <opencv2/ml.hpp>
#include <iostream>

using namespace std;
using namespace cv;
using namespace cv::ml;

int main(int, char**)
{
	// Data for visual representation
	int width = 512, height = 512;
	// zeros就是搞成全0矩阵
	Mat image = Mat::zeros(height, width, CV_8UC3);

	// Set up training data
	//! [setup1]
	//设置训练数据
	int labels[4] = { 1, -1, -1, -1 };  // 每个样本点对应的类
	float trainingData[4][2] = { { 501, 10 }, { 255, 10 }, { 501, 255 }, { 10, 501 } };
	//! [setup1]
	//! [setup2]
	// 将训练数据存入浮点型Mat中
	Mat trainingDataMat(4, 2, CV_32FC1, trainingData);
	// 使用OpenCV里面的机器学习算法时，要保证给的labelData的数据格式为”有符号的整型数”，
	// ”CV_32FC1”就是错误的一种形式，应该使用”CV_32SC1”
	Mat labelsMat(4, 1, CV_32SC1, labels);
	//! [setup2]


	// Train the SVM
	//! [init]
	// 这里的svm是一个指针
	Ptr<SVM> svm = SVM::create();
	svm->setType(SVM::C_SVC);  // 文本选择
	svm->setKernel(SVM::LINEAR);
	// TermCriteria是一个结构体，包括终止的类型，迭代次数或者元素数量，精度。
	svm->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, 100, 1e-6));
	//! [init]
	//! [train],ROW_SAMPLE指每次训练样本的一行
	// 因此总的意思是每次训练样本的一行，而这一行是哪一类是由labelsMat决定的
	svm->train(trainingDataMat, ROW_SAMPLE, labelsMat);
	//! [train]

	// Show the decision regions given by the SVM
	//! [show]
	Vec3b green(0, 255, 0), blue(255, 0, 0);
	for (int i = 0; i < image.rows; ++i)
		for (int j = 0; j < image.cols; ++j)
		{
			Mat sampleMat = (Mat_<float>(1, 2) << j, i);
			float response = svm->predict(sampleMat);

			if (response == 1)
				image.at<Vec3b>(i, j) = green;
			else if (response == -1)
				image.at<Vec3b>(i, j) = blue;
		}
	//! [show]

	// Show the training data
	//! [show_data]
	int thickness = -1;     // 实心圆
	int lineType = 8;
	circle(image, Point(501, 10), 5, Scalar(0, 0, 0), thickness, lineType);
	circle(image, Point(255, 10), 5, Scalar(255, 255, 255), thickness, lineType);
	circle(image, Point(501, 255), 5, Scalar(255, 255, 255), thickness, lineType);
	circle(image, Point(10, 501), 5, Scalar(255, 255, 255), thickness, lineType);
	//! [show_data]

	// Show support vectors
	//! [show_vectors]
	thickness = 2;    // 线宽为2的圆
	lineType = 8;
	// 输出所有的支持向量
	//int c = svm->getVarCount();     // 获得支持向量的维数
	//cout << c << endl;
	Mat sv = svm->getSupportVectors();
	//cout << sv.cols << "-" << sv.rows << endl;

	// 训练出来的SVM classifier的support vector是以Mat类型进行存储，
	//其rows是根据svm的参数而变化的，如果是二分类分类器，应该是1*n的大小。
	// 这里一行存的坐标。
	for (int i = 0; i < sv.rows ; ++i)
	{
		const float* v = sv.ptr<float>(i);
		// 6是圆的半径，(128,128,128)是灰色
		circle(image, Point((int)v[0], (int)v[1]), 10, Scalar(128, 128, 128), thickness, lineType);
	}
	//! [show_vectors]

	imwrite("result.png", image);        // save the image

	imshow("SVM Simple Example", image); // show it to the user
	waitKey(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
#include <opencv2/core.hpp>
#include <opencv2/imgproc.hpp>
#include "opencv2/imgcodecs.hpp"
#include <opencv2/highgui.hpp>
#include <opencv2/ml.hpp>
#include <iostream>

using namespace std;
using namespace cv;
using namespace cv::ml;

int main()
{
	Mat srcImg = imread("test.jpg");
	Mat desImg = srcImg.clone();
	imshow("原图", srcImg);
	// 选取目标区域和背景区域
	Mat BackImg = srcImg(Rect(199, 0, 30, 30));
	Mat ForeImg = srcImg(Rect(38, 95, 30, 30));

	// 初始化训练数据
	Mat trainingDataMat = ForeImg.clone().reshape(1, ForeImg.cols*ForeImg.rows);
	//在这里直接存入背景像素点，或者像下边一个一个点存入也可以
	trainingDataMat.push_back(BackImg.clone().reshape(1,BackImg.cols*BackImg.rows));
	trainingDataMat.convertTo(trainingDataMat, CV_32FC1);

	// 初始化标签，分别给两种标签辅助，虽然这里memset已经全部初始化为1了，可是这里的1是浮点数
	int *labels = new int[ForeImg.cols*ForeImg.rows + BackImg.cols*BackImg.rows];
	memset(labels, 1, sizeof(int)*(ForeImg.cols*ForeImg.rows + BackImg.cols*BackImg.rows));
	for (int i = 0; i < ForeImg.rows; ++i)
		for (int j = 0; j < ForeImg.cols; ++j){
			labels[i*ForeImg.cols + j] = 1;
		}
	for (int h = 0; h<BackImg.rows; h++)
	{
		for (int w = 0; w<BackImg.cols; w++)
		{
			labels[ForeImg.cols*ForeImg.rows + h*BackImg.cols + w] = -1;
		}
	}
	Mat labelsMat = Mat(ForeImg.cols*ForeImg.rows + BackImg.cols*BackImg.rows, 1, CV_32SC1, labels);

	// 可以将数据写入文件来检查是否正确
	//FileStorage fs("data.xml", FileStorage::WRITE);
	//fs << "traindata" << trainingDataMat << "labels" << labelsMat;
	
	// Train the SVM
	Ptr<SVM> svm = SVM::create();
	svm->setType(SVM::C_SVC);
	svm->setKernel(SVM::RBF);
	svm->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, (int)1e5, 1e-6));
	svm->train(trainingDataMat, ROW_SAMPLE, labelsMat);

	// 开始分类
	Vec3b black(0, 0, 0), white(255, 255, 255);
	for (int i = 0; i < desImg.rows; ++i)
	{
		uchar* p_sample = desImg.ptr<uchar>(i);
		for (int j = 0; j < desImg.cols; ++j)
		{
			Mat sampleMat(1, 3, CV_32FC1);
			sampleMat.at<float>(0, 0) = p_sample[3 * j + 0];
			sampleMat.at<float>(0, 1) = p_sample[3 * j + 1];
			sampleMat.at<float>(0, 2) = p_sample[3 * j + 2];

			float response = svm->predict(sampleMat);

			if (response == 1)
				desImg.at<Vec3b>(i, j) = white;
			else if (response == -1)
				desImg.at<Vec3b>(i, j) = black;
		}
	}

	imwrite("result.jpg", desImg);
	imshow("resImg", desImg);
	
	waitKey(0);
	return 0;
}