在opencv3中实现机器学习之：利用正态贝叶斯分类

opencv3.0版本中，实现正态贝叶斯分类器（Normal Bayes Classifier）分类实例

#include "stdafx.h"
#include "opencv2/opencv.hpp"
using namespace cv;
using namespace cv::ml;

int main(int, char**)
{
    int width = , height = ;
    Mat image = Mat::zeros(height, width, CV_8UC3);  //创建窗口可视化

    // 设置训练数据
    int labels[] = { , -, , ,-,,-,,-,- };
    Mat labelsMat(, , CV_32SC1, labels);

    float trainingData[][] = { { ,  }, { ,  }, { ,  }, { ,  }, { ,  },
    { ,  }, { ,  } , { ,  } , { ,  } , { ,  } };
    Mat trainingDataMat(, , CV_32FC1, trainingData);

    // 创建贝叶斯分类器
    Ptr<NormalBayesClassifier> model=NormalBayesClassifier::create();

    // 设置训练数据
    Ptr<TrainData> tData =TrainData::create(trainingDataMat, ROW_SAMPLE, labelsMat);

    //训练分类器
    model->train(tData);

    Vec3b green(, , ), blue(, , );
    // Show the decision regions given by the SVM
    for (int i = ; i < image.rows; ++i)
    for (int j = ; j < image.cols; ++j)
    {
        Mat sampleMat = (Mat_<float>(, ) << j, i);  //生成测试数据
        float response = model->predict(sampleMat);  //进行预测，返回1或-1

        if (response == )
            image.at<Vec3b>(i, j) = green;
        else if (response == -)
            image.at<Vec3b>(i, j) = blue;
    }

    // 显示训练数据
    int thickness = -;
    int lineType = ;
    Scalar c1 = Scalar::all(); //标记为1的显示成黑点
    Scalar c2 = Scalar::all(); //标记成-1的显示成白点
    //绘图时，先宽后高，对应先列后行
    for (int i = ; i < labelsMat.rows; i++)
    {
        const float* v = trainingDataMat.ptr<float>(i); //取出每行的头指针
        Point pt = Point((int)v[], (int)v[]);
        if (labels[i] == )
            circle(image, pt, , c1, thickness, lineType);
        else
            circle(image, pt, , c2, thickness, lineType);

    }

    imshow("normal Bayessian classifier Simple Example", image); // show it to the user
    waitKey();

}

如果将数据换成是图片的像素值 ，则可实现图片的分类。