SVM《一、硬分类、目标函数推导、二分类实验》

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

using namespace cv;
using namespace cv::ml;

#include<iostream>
using namespace std;


void PrintMat(const Mat& mat)
{
    for (int j = 0; j < mat.rows; j++)
    {
        for (int i = 0; i < mat.cols; i++)
        {
            cout << mat.ptr<float>(j)[i] << " ";
        }
        cout << endl;
    }
    cout << endl;
}

int main(int, char**)
{
    // Data for visual representation
    int width = 512, height = 512;
    Mat image = Mat::zeros(height, width, CV_8UC3);

    //<1>训练数据
    int labels[3] = { 1, 1, -1 };
    float trainingData[3][2] = { { 30, 30 },{ 40, 40 },{10, 10 }};
    Mat trainingDataMat(3, 2, CV_32FC1, trainingData);
    Mat labelsMat(3, 1, CV_32SC1, labels);

    //SVM初始化并开始训练
    Ptr<SVM> svm = SVM::create();
    svm->setType(SVM::C_SVC);
    svm->setKernel(SVM::LINEAR);
    svm->setTermCriteria(TermCriteria(TermCriteria::MAX_ITER, 100, 1e-6));
    svm->train(trainingDataMat, ROW_SAMPLE, labelsMat);

    //预测 验证样本
    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;
        }

    // 显示训练数据
    int thickness = -1;
    int lineType = 8;
    circle(image, Point(30, 30), 5, Scalar(0, 0, 0), thickness, lineType);
    circle(image, Point(40, 40), 5, Scalar(0, 0, 0), thickness, lineType);
    circle(image, Point(10, 10), 5, Scalar(255, 255, 255), thickness, lineType);

    //显示支撑向量
    thickness = 2;
    lineType = 8;

    Mat sv = svm->getUncompressedSupportVectors();
    PrintMat(sv);
    Mat sv_ = svm->getSupportVectors();
    PrintMat(sv_);

    for (int i = 0; i < sv.rows; ++i)
    {
        const float* v = sv.ptr<float>(i);
        circle(image, Point((int)v[0], (int)v[1]), 6, Scalar(0, 0, 255), thickness, lineType);
    }
    // imwrite("result.png", image);        // save the image
    imshow("SVM Simple Example", image); // show it to the user
    waitKey(0);

}

如图，白色点是正样本、黑色点是负样本。红色点表示支撑向量上的点。绿色和蓝色是svm->predict的结果。

如图：Mat sv = svm->getUncompressedSupportVectors(); 得出支撑向量上的样本点，打印出来就是(30 ,30) (10, 10)

Mat sv_ = svm->getSupportVectors(); 得出决策面方程，和我们上述计算的一致(这里点都放大了10倍)，不知为什么没有超平面的截距

 参考：https://blog.csdn.net/chaipp0607/article/details/73662441