程序片段----opencv cv::ml::KNearest knn 20170904

// opencv3 knn 的实例
// 样本是随机数生成的，不需要额外数据集。
// knn  : k 是要设定的参数，意义是：将待测样本X最近的k个点进行比较，A类型的点最多，那么认为待测样本X是A类型。
// 环境 : opencv3.0.0 \ vs2012 32 bits \ win7
// 环境搭建：
//	# 1. 新建工程，opencv基本配置
//  # 2. 将 opencv\source\module\ml\src 中的 knearest.cpp 复制到新建的工程目录下，文件名改为 knearest.hpp 
//  # 3. 包含该头文件
//  # 4. mian函数修改如下
//  # 5. 编译运行。
// 



#include 
#include 
#include 
#include 
#include 

using namespace cv::ml;

int main( )
{
    const int K = 10;
    int i, j, k, accuracy;
    float response;
    int train_sample_count = 100;
    cv::RNG rng_state(-1);
    cv::Mat trainData(train_sample_count,2,CV_32FC1);
    cv::Mat trainClasses(train_sample_count,1,CV_32FC1); /// labels
    cv::Mat img(cv::Size(500,500),CV_8UC3,cv::Scalar::all (0));
    float _sample[2];
    cv::Mat sample(1,2,CV_32FC1,_sample); /// just 1 sample

    cv::Mat trainData1, trainData2, trainClasses1, trainClasses2;

    // form the training samples
    trainData1 = trainData.rowRange (0,train_sample_count/2);
    rng_state.fill (trainData1,CV_RAND_NORMAL,cv::Scalar(200,200),cv::Scalar(50,50));

    trainData2 = trainData.rowRange (train_sample_count/2,train_sample_count);
    rng_state.fill (trainData2,CV_RAND_NORMAL,cv::Scalar(300,300),cv::Scalar(50,50));

    trainClasses1 = trainClasses.rowRange (0,train_sample_count/2);
    trainClasses1.setTo (1);

    trainClasses2 = trainClasses.rowRange (train_sample_count/2,train_sample_count);
    trainClasses2.setTo (2);

    // learn classifier
	//// cv::ml::KNearest knn( trainData, trainClasses, cv::Mat(), false, K );
	cv::Ptr knn = KNearest::create();
	knn->setDefaultK(5);
	knn->setIsClassifier(true);
	cv::Ptr tData = TrainData::create(trainData, ROW_SAMPLE, trainClasses);
	knn->train(tData);

    cv::Mat nearests( 1, K, CV_32FC1); //// closet k points
	

    for( i = 0; i < img.rows; i++ ) //// 将图中各点作为样本。红是1类型置信值高的，绿是2类型置信值高的，棕色是不确定
    {
        for( j = 0; j < img.cols; j++ )
        {
            sample.at(0,0) = (float)j;
            sample.at(0,1) = (float)i;
			cv::Mat result(sample.size(), CV_32FC1);

            // estimate the response and get the neighbors' labels
			/// response = knn->findNearest(sample,K,0,0,&nearests,0);
			response = knn->findNearest(sample, K, result, nearests);

            // compute the number of neighbors representing the majority
            for( k = 0, accuracy = 0; k < K; k++ )
            {
                if( nearests.at(0,k) == response)
                    accuracy++; /// 最近邻的k个中，该类型的占比
            }
            // highlight the pixel depending on the accuracy (or confidence)
            img.at(i,j) = response == 1 ?
                        (accuracy > 5 ? cv::Vec3b(0,0,180) : cv::Vec3b(0,120,180)) :
                        (accuracy > 5 ? cv::Vec3b(0,180,0) : cv::Vec3b(0,120,120));
        }
    }

    // display the original training samples
    for( i = 0; i < train_sample_count/2; i++ ) ///// 圈 是样本
    {
        cv::Point pt;
        pt.x = cvRound(trainData1.at(i,0));
        pt.y = cvRound(trainData1.at(i,1));
        cv::circle (img,pt,2,cv::Scalar(0,0,255),1,CV_FILLED);
        pt.x = cvRound(trainData2.at(i,0));
        pt.y = cvRound(trainData2.at(i,1));
        cv::circle (img,pt,2,cv::Scalar(0,255,0),1,CV_FILLED);
    }

    cv::namedWindow( "classifier result", 1 );
    cv::imshow( "classifier result", img );
    cv::waitKey(0);

    return 0;
}