低鲁棒车流分析

具体问题交流请联系：clr_mv@163.com
更多文章请关注微信公众号：机器视觉专业论坛

本例中采用halcon的图像序列来简单分析车辆的流动，分析方法较为简单，通过
前后两幅图像的对比来进行，因此鲁棒性不强。
本例中学习到的知识主要有：
1.Videocapture类读入图像序列时对于PNG格式的图像无效。
2.在进行图像复制时不可以简单的使用“=”，需要copyto或者clone。

#include <iostream>     
#include <opencv2/opencv.hpp>  
#include <opencv2/highgui/highgui.hpp>
#include <opencv2/core/core.hpp>


using namespace cv;
using namespace std;


int main(int argc, char** argv)
{

    string file ;
    RNG rng(12345);
    Mat intimage, image, diff, offset, imagecolor, intimagecolor,tim;
    file = "G:\\OpencvStudy\\Test3\\data\\xing\\init.png";
    intimagecolor = cv::imread(file);
    if (intimagecolor.empty()) 
        return -1;
    cvtColor(intimagecolor, intimage, CV_RGB2GRAY);
    for (int i = 0; i < 586; i++)

    {
        string s;
        char t[256];
        sprintf_s(t, "%03d", i);
        s = t;//int 转化为String

        file = "G:\\OpencvStudy\\Test3\\data\\xing\\xing" + s + ".png";
        imagecolor = cv::imread(file);
        if (imagecolor.empty())
            break;

        cvtColor(imagecolor, image, CV_RGB2GRAY);//转化为灰度图，便于后续分析
        GaussianBlur(image, image, Size(3, 3), 0.4, 0.4);
        GaussianBlur(intimage, intimage, Size(3, 3), 0.4, 0.4);//对前后两幅图滤波

        absdiff(image, intimage, diff);//求取前后两幅图不同的区域
        intimage = image.clone();//保留初始图像，为下一帧图像分析准备
        addWeighted(diff, 0.5, diff, 0.5, -35,diff);//参见halcon 的函数dyn_threshold
        imshow("track", diff);

        Mat threshold_output;
        vector<vector<Point> > contours;
        vector<Vec4i> hierarchy;

        /// 使用Threshold对图像二值化
        threshold(diff, threshold_output, 10, 255, THRESH_BINARY);

        cv::Mat element5(2, 2, CV_8U, cv::Scalar(1));
        cv::Mat open;
        cv::morphologyEx(diff, open, cv::MORPH_OPEN, element5);

        cv::Mat element15(8, 8, CV_8U, cv::Scalar(1));
        cv::Mat close;
        cv::morphologyEx(open, close, cv::MORPH_CLOSE, element15);

        /// 找到轮廓
        findContours(close, contours, hierarchy, CV_RETR_TREE, CV_CHAIN_APPROX_SIMPLE, Point(0, 0));

        /// 多边形逼近轮廓 + 获取矩形和圆形边界框
        vector<vector<Point> > contours_poly(contours.size());
        vector<Rect> boundRect(contours.size());
        vector<Point2f>center(contours.size());
        vector<float>radius(contours.size());

        for (int i = 0; i < contours.size(); i++)
        {
            approxPolyDP(Mat(contours[i]), contours_poly[i], 3, true);
            boundRect[i] = boundingRect(Mat(contours_poly[i]));
            minEnclosingCircle(contours_poly[i], center[i], radius[i]);
        }


        /// 画多边形轮廓 + 包围的矩形框 + 圆形框
        Mat drawing = Mat::zeros(threshold_output.size(), CV_8UC3);
        for (int i = 0; i< contours.size(); i++)
        {
            Scalar color = Scalar(rng.uniform(0, 255), rng.uniform(0, 255), rng.uniform(0, 255));
            //drawContours(drawing, contours_poly, i, color, 1, 8, vector<Vec4i>(), 0, Point());
            rectangle(drawing, boundRect[i].tl(), boundRect[i].br(), color, 2, 8, 0);
            //circle(drawing, center[i], (int)radius[i], color, 2, 8, 0);
        }
        imshow("Drawing", drawing);
        add(imagecolor, drawing, imagecolor);
        imshow("Image sequence", imagecolor);
        waitKey(5);//每隔5毫秒再读下一幅图像     
    }



    return 0;
}

本文分析思路较为简单，主要是为了熟悉Opencv相应算法。