平面中点间距小于阈值的异常点探测

问题描述：
1.随机生成平面中的二维点；
2.探测点间距小于距离阈值的异常点；
3.显示并输出结果。

#include "stdafx.h"
#include "iostream"
#include "string"
#include "fstream"
#include "sstream"
#include "stdlib.h"
#include "time.h"
#include "vector"
#include <opencv\highgui.h>
#include <opencv\cv.h>
#include <opencv\ml.h>
#include <opencv\cxcore.h>

#include <opencv2\core\core.hpp>
#include <opencv2\highgui\highgui.hpp>
#include "opencv2\nonfree\features2d.hpp"
#include <opencv2\flann\miniflann.hpp>
#include <opencv2\flann\flann.hpp>
#include <opencv2\flann\kdtree_index.h>

using namespace std;
using namespace cv;
int _tmain(int argc, _TCHAR* argv[])
{
    //*****①生成随机点（0-1000之内）*****//
    vector<Point2i> random_point;
    Point2i temp_p;
    int number_random_point=1000;
    cv::Mat_<float> source_p(number_random_point, 2);
    srand((unsigned)time(NULL));
    for (int i = 0; i < number_random_point; i++)
    {
        temp_p.x = (int)1000 * rand() / (RAND_MAX + 1);
        temp_p.y = (int)1000 * rand() / (RAND_MAX + 1);//get the random point within （0,1000）
        random_point.push_back(temp_p);
        cv::Mat this_row(1,2,CV_16U);
        this_row.at<ushort>(0, 0) = temp_p.x;
        this_row.at<ushort>(0, 1) = temp_p.y;
        source_p.at<float>(i, 0) = (float)temp_p.x;
        source_p.at<float>(i, 1) = (float)temp_p.y;

    }
    cout << source_p << endl;
    //*****②找出小于距离阈值d的异常点*****//
    //index construction
    vector<int> outlier_id; 


    cv::flann::Index flann_index(source_p, cv::flann::KDTreeIndexParams(1));
    unsigned int max_searched_neighbours = 6;   
    float threshold_d = 30.0;
    for (int i = 0; i < number_random_point; i++)
    {
        Mat_<float> my_query(1, 2);
        my_query.at<float>(0, 0) = source_p.at<float>(i, 0); 
        my_query.at<float>(0, 1) = source_p.at<float>(i, 1); 
        cout << my_query << endl;
        Mat index_p, index_distance;        
        flann_index.radiusSearch(my_query, index_p, index_distance, threshold_d,max_searched_neighbours);
        cout << index_p << endl;
        cout << index_distance << endl;
        if (!index_p.empty())
        {
            for (int j = 0; j < index_p.cols; j++)
            {
                if (index_p.at<int>(0, j)<0 || index_p.at<int>(0, j)>number_random_point)
                {
                    continue;//如果找到的存贮的点号不在（0，num_randon_point）中
                }
                if (index_distance.at<float>(0, j)<= 0 || index_distance.at<float>(0, j)>threshold_d)
                {
                    continue;//如果距离不在阈值范围之内
                }
                if (find(outlier_id.begin(),outlier_id.end(),index_p.at<int>(0,j))==outlier_id.end())
                {
                    outlier_id.push_back(index_p.at<int>(0,j));
                }
            }
        }
    }

    //******③输出异常点********//
    Mat dst_img(6000, 6000, CV_AA, Scalar(255, 255, 255));
    float circle_radius = 3.0;
    vector<Point2i>::iterator it1 = random_point.begin();
    Point2i temp_center;
    while (it1!=random_point.end())
    {
        temp_center.x = it1->x * 4;
        temp_center.y = it1->y * 4;
        circle(dst_img, temp_center, circle_radius, Scalar(0, 255, 0),-1);
        it1++;
    }
    imshow("original_points", dst_img);
    waitKey(0);
    for (auto it = outlier_id.begin(); it != outlier_id.end(); it++)
    {
        Point2i circle_center = random_point[*it];
        circle_center.x = circle_center.x * 4;
        circle_center.y = circle_center.y * 4;
        circle(dst_img, circle_center, circle_radius, Scalar(0, 0, 255), -1);
    }

    imshow("found_outliers", dst_img);
    waitKey(0);
    string output_filename = "outlier.bmp";
    imwrite(output_filename, dst_img);
    return 0;
}