9.4 【OpenCV】 基于 ransac 算法的 sift 特征匹配程序（开发环境为OpenCV2.3.1+VS2010）

原地址：http://blog.csdn.net/zouwen198317/article/details/38493777

[cpp] view plaincopy
因为前一阵子忙于自己的毕设，所以就没有及时更新日志，今天正好没其他事儿，所以，我就把图像拼接程序写上来了。。。欢迎大家的阅读以及批评和指正。  
下面的程序是基于opencv2.3.1+vs2010的搭建的环境下编程的。。。  
首先对两个usb通用摄像头进行了标定。然后进行图像拼接，最后进行测距。  
这不是最终版，因为最终版是我的论文内容。  
所以，要过一阵子才能写上来，因为现在写上来我的论文可能被互联网的查重率坑爹的。。。  
呵呵，请大家见谅。。。不过，很多重点代码都是完好无损的。。。。  
  
#include <math.h>  
#include <ctime>  
#include <cv.h>  
#include <highgui.h>  
#include <features2d/features2d.hpp>  
#include <cvaux.h>  
#include <string>  
#include <iostream>  
#include <fstream>  
using namespace std;  
using namespace cv;  
  
  
void main()  
{  
ofstream fout("Result.txt");  
  
  
float f_left[2],f_right[2];  
IplImage * show; //RePlay图像指针  
    cvNamedWindow("RePlay",1);  
    int number_image_copy=7; //复制图像帧数  
    CvSize board_size=cvSize(9,6); //标定板角点数  
    CvSize2D32f square_size=cvSize2D32f(10,10);  //假设我的每个标定方格长宽都是1.82厘米  
    float square_length=square_size.width; //方格长度  
    float square_height=square_size.height;  //方格高度  
    int board_width=board_size.width; //每行角点数  
    int board_height=board_size.height; //每列角点数  
    int total_per_image=board_width*board_height; //每张图片角点总数  
    int count; //存储每帧图像中实际识别的角点数  
    int found; //识别标定板角点的标志位  
    int step; //存储步长，step=successes*total_per_image;  
    int successes=0; //存储成功找到标定板上所有角点的图像帧数  
    int a=1; //临时变量，表示在操作第a帧图像  
const int NImages = 7;//图片总数  
  
  
CvMat *rotation_vectors;  
    CvMat *translation_vectors;  
    CvPoint2D32f * image_points_buf = new CvPoint2D32f[total_per_image]; //存储角点图像坐标的数组  
    CvMat * image_points=cvCreateMat(NImages*total_per_image,2,CV_32FC1);//存储角点的图像坐标的矩阵  
    CvMat * object_points=cvCreateMat(NImages*total_per_image,3,CV_32FC1);//存储角点的三维坐标的矩阵  
    CvMat * point_counts=cvCreateMat(NImages,1,CV_32SC1);//存储每帧图像的识别的角点数  
    CvMat * intrinsic_matrix=cvCreateMat(3,3,CV_32FC1);//内参数矩阵  
    CvMat * distortion_coeffs=cvCreateMat(5,1,CV_32FC1);//畸变系数  
    rotation_vectors = cvCreateMat(NImages,3,CV_32FC1);//旋转矩阵  
    translation_vectors = cvCreateMat(NImages,3,CV_32FC1);//平移矩阵  
ifstream fin("calibdata1.txt"); /* 定标所用图像文件的路径 */  
  
  
    while(a<=number_image_copy)  
    {  
        //sprintf_s (filename,"%d.jpg",a);  
string filename;  
getline(fin,filename);  
        show=cvLoadImage(filename.c_str(),1);  
        //寻找棋盘图的内角点位置  
        found=cvFindChessboardCorners(show,board_size,image_points_buf,&count,  
        CV_CALIB_CB_ADAPTIVE_THRESH|CV_CALIB_CB_FILTER_QUADS);  
  
  
        if (found==0)  
        { //如果没找到标定板角点时  
            cout<<"第"<<a<<"帧图片无法找到棋盘格所有角点!\n\n";  
fout<<"第"<<a<<"帧图片无法找到棋盘格所有角点!\n\n";  
            cvNamedWindow("RePlay",1);  
            cvShowImage("RePlay",show);  
            cvWaitKey(2000);  
  
  
        }  
        else  
        { //找到标定板角点时  
            cout<<"第"<<a<<"帧图像成功获得"<<count<<"个角点...\n";  
fout<<"第"<<a<<"帧图像成功获得"<<count<<"个角点...\n";  
            cvNamedWindow("RePlay",1);  
            IplImage * gray_image= cvCreateImage(cvGetSize(show),8,1);  
            cvCvtColor(show,gray_image,CV_BGR2GRAY);  
            cout<<"获取源图像灰度图过程完成...\n";  
fout<<"获取源图像灰度图过程完成...\n";  
  
  
            cvFindCornerSubPix(gray_image,image_points_buf,count,cvSize(11,11),cvSize(-1,-1),  
            cvTermCriteria(CV_TERMCRIT_EPS+CV_TERMCRIT_ITER,30,0.1));  
            cout<<"灰度图亚像素化过程完成...\n";  
fout<<"灰度图亚像素化过程完成...\n";  
            cvDrawChessboardCorners(show,board_size,image_points_buf,count,found);  
            cout<<"在源图像上绘制角点过程完成...\n\n";  
fout<<"在源图像上绘制角点过程完成...\n\n";  
            cvShowImage("RePlay",show);  
            cvWaitKey(500);  
        }  
  
  
        if(total_per_image==count)  
        {  
            step=successes*total_per_image; //计算存储相应坐标数据的步长  
            for(int i=step,j=0;j<total_per_image;++i,++j)  
            {  
                CV_MAT_ELEM(*image_points, float,i,0)=image_points_buf[j].x;  
                CV_MAT_ELEM(*image_points, float,i,1)=image_points_buf[j].y;  
                CV_MAT_ELEM(*object_points,float,i,0)=(float)((j/board_width)*square_length);  
                CV_MAT_ELEM(*object_points,float,i,1)=(float)((j%board_width)*square_height);  
                CV_MAT_ELEM(*object_points,float,i,2)=0.0f;  
            }  
            CV_MAT_ELEM(*point_counts,int,successes,0)=total_per_image;  
            successes++;  
        }  
cout<<"hahahahh======"<<a<<endl;  
fout<<"hahahahh======"<<a<<endl;  
        a++;  
    }  
  
  
    cvReleaseImage(&show);  
    cvDestroyWindow("RePlay");  
  
  
  
  
    cout<<"*********************************************\n";  
fout<<"*********************************************\n";  
    cout<<NImages<<"帧图片中，标定成功的图片为"<<successes<<"帧...\n";  
fout<<NImages<<"帧图片中，标定成功的图片为"<<successes<<"帧...\n";  
    cout<<NImages<<"帧图片中，标定失败的图片为"<<NImages-successes<<"帧...\n\n";  
fout<<NImages<<"帧图片中，标定失败的图片为"<<NImages-successes<<"帧...\n\n";  
    cout<<"*********************************************\n\n";  
fout<<"*********************************************\n\n";  
  
  
    cout<<"按任意键开始计算摄像机内参数...\n\n";  
fout<<"按任意键开始计算摄像机内参数...\n\n";  
  
  
  
  
    /*CvCapture* capture1; 
    capture1 = cvCreateCameraCapture(0);*/  
    IplImage * show_colie;  
    show_colie = cvLoadImage("F:\\graduatelunwen\\opencvprojects\\june\\opecv_ransac_sift_cameraCalibration\\dinggao_wanchengban_sift_ransac\\left_43.jpg",1);  
  
  
  
  
    CvMat * object_points2 = cvCreateMat(successes*total_per_image,3,CV_32FC1);  
    CvMat * image_points2  = cvCreateMat(successes*total_per_image,2,CV_32FC1);  
    CvMat * point_counts2  = cvCreateMat(successes,1,CV_32SC1);  
  
  
  
  
    for(int i=0;i<successes*total_per_image;++i)  
    {  
        CV_MAT_ELEM(*image_points2, float,i,0)=CV_MAT_ELEM(*image_points, float,i,0);  
        CV_MAT_ELEM(*image_points2, float,i,1)=CV_MAT_ELEM(*image_points, float,i,1);  
        CV_MAT_ELEM(*object_points2,float,i,0)=CV_MAT_ELEM(*object_points,float,i,0);  
        CV_MAT_ELEM(*object_points2,float,i,1)=CV_MAT_ELEM(*object_points,float,i,1);  
        CV_MAT_ELEM(*object_points2,float,i,2)=CV_MAT_ELEM(*object_points,float,i,2);  
    }  
  
  
    for(int i=0;i<successes;++i)  
    {  
        CV_MAT_ELEM(*point_counts2,int,i,0) = CV_MAT_ELEM(*point_counts,int,i,0);  
    }  
  
  
  
  
    cvReleaseMat(&object_points);  
    cvReleaseMat(&image_points);  
    cvReleaseMat(&point_counts);  
  
  
  
  
    CV_MAT_ELEM(*intrinsic_matrix,float,0,0)=1.0f;  
    CV_MAT_ELEM(*intrinsic_matrix,float,1,1)=1.0f;  
  
  
  
    cvCalibrateCamera2(object_points2,image_points2,point_counts2,cvGetSize(show_colie),  
    intrinsic_matrix,distortion_coeffs,rotation_vectors,translation_vectors,0);  
  
  
    cout<<"摄像机内参数矩阵为：\n";  
fout<<"摄像机内参数矩阵为：\n";  
    cout<<CV_MAT_ELEM(*intrinsic_matrix,float,0,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,0,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,0,2)  
    <<"\n\n";  
    cout<<CV_MAT_ELEM(*intrinsic_matrix,float,1,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,1,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,1,2)  
    <<"\n\n";  
    cout<<CV_MAT_ELEM(*intrinsic_matrix,float,2,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,2,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,2,2)  
    <<"\n\n";  
  
  
fout<<CV_MAT_ELEM(*intrinsic_matrix,float,0,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,0,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,0,2)  
    <<"\n\n";  
    fout<<CV_MAT_ELEM(*intrinsic_matrix,float,1,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,1,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,1,2)  
    <<"\n\n";  
    fout<<CV_MAT_ELEM(*intrinsic_matrix,float,2,0)<<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,2,1)  
    <<" "<<CV_MAT_ELEM(*intrinsic_matrix,float,2,2)  
    <<"\n\n";  
  
  
f_left[0]=CV_MAT_ELEM(*intrinsic_matrix,float,0,0);  
f_left[1]=CV_MAT_ELEM(*intrinsic_matrix,float,1,1);  
  
  
  
  
    cout<<"畸变系数矩阵为：\n";  
    cout<<CV_MAT_ELEM(*distortion_coeffs,float,0,0)<<" "<<CV_MAT_ELEM(*distortion_coeffs,float,1,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,2,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,3,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,4,0)  
    <<"\n\n";  
  
  
fout<<"畸变系数矩阵为：\n";  
    fout<<CV_MAT_ELEM(*distortion_coeffs,float,0,0)<<" "<<CV_MAT_ELEM(*distortion_coeffs,float,1,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,2,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,3,0)  
    <<" "<<CV_MAT_ELEM(*distortion_coeffs,float,4,0)  
    <<"\n\n";  
  
  
    cvSave("Intrinsics.xml",intrinsic_matrix);  
    cvSave("Distortion.xml",distortion_coeffs);  
  
  
    cout<<"摄像机矩阵、畸变系数向量已经分别存储在名为Intrinsics.xml、Distortion.xml文档中\n\n";  
    fout<<"摄像机矩阵、畸变系数向量已经分别存储在名为Intrinsics.xml、Distortion.xml文档中\n\n";  
  
  
for(int ii = 0; ii < NImages; ++ii)  
{ float tranv[3] = {0.0};  
float rotv[3] = {0.0};  
  
  
for ( int i = 0; i < 3; i++)  
{  
tranv[i] = ((float*)(translation_vectors->data.ptr+ii*translation_vectors->step))[i];  
rotv[i] = ((float*)(rotation_vectors->data.ptr+rotation_vectors->step))[i];  
}  
  
cout << "第" << ii+1 << "幅图的外参数" << endl;  
fout << "第" << ii+1 << "幅图的外参数" << endl;  
printf("ROTATION VECTOR 旋转向量 : \n");  
printf("[ %6.4f %6.4f %6.4f ] \n", rotv[0], rotv[1], rotv[2]);  
printf("TRANSLATION VECTOR 平移向量: \n");  
printf("[ %6.4f %6.4f %6.4f ] \n", tranv[0], tranv[1], tranv[2]);  
printf("-----------------------------------------\n");  
  
  
fout<<"ROTATION VECTOR 旋转向量 : \n"<<endl;  
fout<< rotv[0]<<"  "<< rotv[1]<<"  "<< rotv[2]<<endl;  
fout<<"TRANSLATION VECTOR 平移向量: \n"<<endl;  
fout<< tranv[0]<<"  "<< tranv[1]<< "  "<<tranv[2]<<endl;  
fout<<"-----------------------------------------\n"<<endl;  
}  
    CvMat * intrinsic=(CvMat *)cvLoad("Intrinsics.xml");  
    CvMat * distortion=(CvMat *)cvLoad("Distortion.xml");  
  
  
    IplImage * mapx=cvCreateImage(cvGetSize(show_colie),IPL_DEPTH_32F,1);  
    IplImage * mapy=cvCreateImage(cvGetSize(show_colie),IPL_DEPTH_32F,1);  
  
  
    cvInitUndistortMap(intrinsic,distortion,mapx,mapy);  
  
  
    cvNamedWindow("原始图像",1);  
    cvNamedWindow("非畸变图像",1);  
  
  
   // cout<<"按‘E’键退出显示...\n\n";  
  
  
    /*while(show_colie) 
    {*/  
        IplImage * clone=cvCloneImage(show_colie);  
        cvShowImage("原始图像",show_colie);  
        cvRemap(clone,show_colie,mapx,mapy);  
        cvReleaseImage(&clone);  
        cvShowImage("非畸变图像",show_colie);  
cvWaitKey(500);  
        /*if(cvWaitKey(10)=='e') 
        { 
            break; 
        }*/  
  
  
        /*show_colie=cvQueryFrame(capture1); 
    }*/  
  
  
  
  
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////  
///////////                                                        //////////////////////////////////////////////  
//////////                右摄像机标定                       //////////////////////////////////////////////                                  
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////  
  
    IplImage * showR; //RePlay图像指针  
    cvNamedWindow("RePlayR",1);  
    int number_image_copyR=7; //复制图像帧数  
    CvSize board_sizeR=cvSize(9,6); //标定板角点数  
    CvSize2D32f square_sizeR=cvSize2D32f(10,10);  //假设我的每个标定方格长宽都是1.82厘米  
    float square_lengthR=square_sizeR.width;  //方格长度  
    float square_heightR=square_sizeR.height;  //方格高度  
    int board_widthR=board_sizeR.width; //每行角点数  
    int board_heightR=board_sizeR.height; //每列角点数  
    int total_per_imageR=board_widthR*board_heightR; //每张图片角点总数  
    int countR; //存储每帧图像中实际识别的角点数  
    int foundR; //识别标定板角点的标志位  
    int stepR; //存储步长，stepR=successesR*total_per_imageR;  
    int successesR=0; //存储成功找到标定板上所有角点的图像帧数  
    int aR=1; //临时变量，表示在操作第a帧图像  
const int NImagesR = 7;//图片总数  
  
  
CvMat *rotation_vectorsR;  
    CvMat *translation_vectorsR;  
    CvPoint2D32f * image_pointsR_bufR = new CvPoint2D32f[total_per_imageR]; //存储角点图像坐标的数组  
    CvMat * image_pointsR=cvCreateMat(NImagesR*total_per_imageR,2,CV_32FC1);//存储角点的图像坐标的矩阵  
    CvMat * object_pointsR=cvCreateMat(NImagesR*total_per_imageR,3,CV_32FC1);//存储角点的三维坐标的矩阵  
    CvMat * point_countRsR=cvCreateMat(NImagesR,1,CV_32SC1);//存储每帧图像的识别的角点数  
    CvMat * intrinsicR_matrixR=cvCreateMat(3,3,CV_32FC1);//内参数矩阵  
    CvMat * distortionR_coeffsR=cvCreateMat(5,1,CV_32FC1);//畸变系数  
    rotation_vectorsR = cvCreateMat(NImagesR,3,CV_32FC1);//旋转矩阵  
    translation_vectorsR = cvCreateMat(NImagesR,3,CV_32FC1);//平移矩阵  
ifstream finR("calibdata2.txt"); /* 定标所用图像文件的路径 */  
  
  
    while(aR<=number_image_copyR)  
    {  
        //sprintf_s (filename,"%d.jpg",a);  
string filenameR;  
getline(finR,filenameR);  
        showR=cvLoadImage(filenameR.c_str(),1);  
        //寻找棋盘图的内角点位置  
        foundR=cvFindChessboardCorners(showR,board_sizeR,image_pointsR_bufR,&countR,  
        CV_CALIB_CB_ADAPTIVE_THRESH|CV_CALIB_CB_FILTER_QUADS);  
  
  
        if (foundR==0)  
        { //如果没找到标定板角点时  
            cout<<"第"<<aR<<"帧图片无法找到棋盘格所有角点!\n\n";  
fout<<"第"<<aR<<"帧图片无法找到棋盘格所有角点!\n\n";  
            cvNamedWindow("RePlayR",1);  
            cvShowImage("RePlayR",showR);  
            cvWaitKey(500);  
  
  
        }  
        else  
        { //找到标定板角点时  
            cout<<"第"<<aR<<"帧图像成功获得"<<countR<<"个角点...\n";  
fout<<"第"<<aR<<"帧图像成功获得"<<countR<<"个角点...\n";  
            cvNamedWindow("RePlayR",1);  
            IplImage * gray_imageR= cvCreateImage(cvGetSize(showR),8,1);  
            cvCvtColor(showR,gray_imageR,CV_BGR2GRAY);  
            cout<<"获取源图像灰度图过程完成...\n";  
fout<<"获取源图像灰度图过程完成...\n";  
  
  
            cvFindCornerSubPix(gray_imageR,image_pointsR_bufR,countR,cvSize(11,11),cvSize(-1,-1),  
            cvTermCriteria(CV_TERMCRIT_EPS+CV_TERMCRIT_ITER,30,0.1));  
            cout<<"灰度图亚像素化过程完成...\n";  
fout<<"灰度图亚像素化过程完成...\n";  
            cvDrawChessboardCorners(showR,board_sizeR,image_pointsR_bufR,countR,foundR);  
            cout<<"在源图像上绘制角点过程完成...\n\n";  
fout<<"在源图像上绘制角点过程完成...\n\n";  
            cvShowImage("RePlayR",showR);  
            cvWaitKey(500);  
        }  
  
  
        if(total_per_imageR==countR)  
        {  
            stepR=successesR*total_per_imageR; //计算存储相应坐标数据的步长  
            for(int i=stepR,j=0;j<total_per_imageR;++i,++j)  
            {  
                CV_MAT_ELEM(*image_pointsR, float,i,0)=image_pointsR_bufR[j].x;  
                CV_MAT_ELEM(*image_pointsR, float,i,1)=image_pointsR_bufR[j].y;  
                CV_MAT_ELEM(*object_pointsR,float,i,0)=(float)((j/board_widthR)*square_lengthR);  
                CV_MAT_ELEM(*object_pointsR,float,i,1)=(float)((j%board_widthR)*square_heightR);  
                CV_MAT_ELEM(*object_pointsR,float,i,2)=0.0f;  
            }  
            CV_MAT_ELEM(*point_countRsR,int,successesR,0)=total_per_imageR;  
            successesR++;  
        }  
cout<<"hahahahh======"<<aR<<endl;  
fout<<"hahahahh======"<<aR<<endl;  
        aR++;  
    }  
  
  
    cvReleaseImage(&showR);  
    cvDestroyWindow("RePlayR");  
  
  
  
  
    cout<<"*********************************************\n";  
    cout<<NImagesR<<"帧图片中，标定成功的图片为"<<successesR<<"帧...\n";  
    cout<<NImagesR<<"帧图片中，标定失败的图片为"<<NImagesR-successesR<<"帧...\n\n";  
    cout<<"*********************************************\n\n";  
  
  
    cout<<"按任意键开始计算摄像机内参数...\n\n";  
  
  
fout<<"*********************************************\n";  
    fout<<NImagesR<<"帧图片中，标定成功的图片为"<<successesR<<"帧...\n";  
    fout<<NImagesR<<"帧图片中，标定失败的图片为"<<NImagesR-successesR<<"帧...\n\n";  
    fout<<"*********************************************\n\n";  
  
  
    fout<<"按任意键开始计算摄像机内参数...\n\n";  
  
  
  
  
    /*CvCapture* capture1; 
    capture1 = cvCreateCameraCapture(0);*/  
    IplImage * showR_colieR;  
    showR_colieR = cvLoadImage("F:\\graduatelunwen\\opencvprojects\\june\\opecv_ransac_sift_cameraCalibration\\dinggao_wanchengban_sift_ransac\\right_43.jpg",1);  
  
  
  
  
    CvMat * object_pointsR2R = cvCreateMat(successesR*total_per_imageR,3,CV_32FC1);  
    CvMat * image_pointsR2R  = cvCreateMat(successesR*total_per_imageR,2,CV_32FC1);  
    CvMat * point_countRsR2R  = cvCreateMat(successesR,1,CV_32SC1);  
  
  
  
  
    for(int i=0;i<successesR*total_per_imageR;++i)  
    {  
        CV_MAT_ELEM(*image_pointsR2R, float,i,0)=CV_MAT_ELEM(*image_pointsR, float,i,0);  
        CV_MAT_ELEM(*image_pointsR2R, float,i,1)=CV_MAT_ELEM(*image_pointsR, float,i,1);  
        CV_MAT_ELEM(*object_pointsR2R,float,i,0)=CV_MAT_ELEM(*object_pointsR,float,i,0);  
        CV_MAT_ELEM(*object_pointsR2R,float,i,1)=CV_MAT_ELEM(*object_pointsR,float,i,1);  
        CV_MAT_ELEM(*object_pointsR2R,float,i,2)=CV_MAT_ELEM(*object_pointsR,float,i,2);  
    }  
  
  
    for(int i=0;i<successesR;++i)  
    {  
        CV_MAT_ELEM(*point_countRsR2R,int,i,0) = CV_MAT_ELEM(*point_countRsR,int,i,0);  
    }  
  
  
  
  
    cvReleaseMat(&object_pointsR);  
    cvReleaseMat(&image_pointsR);  
    cvReleaseMat(&point_countRsR);  
  
  
  
  
    CV_MAT_ELEM(*intrinsicR_matrixR,float,0,0)=1.0f;  
    CV_MAT_ELEM(*intrinsicR_matrixR,float,1,1)=1.0f;  
  
  
  
    cvCalibrateCamera2(object_pointsR2R,image_pointsR2R,point_countRsR2R,cvGetSize(showR_colieR),  
    intrinsicR_matrixR,distortionR_coeffsR,rotation_vectorsR,translation_vectorsR,0);  
  
  
    cout<<"摄像机内参数矩阵为：\n";  
    cout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,2)  
    <<"\n\n";  
    cout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,2)  
    <<"\n\n";  
    cout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,2)  
    <<"\n\n";  
  
  
fout<<"摄像机内参数矩阵为：\n";  
    fout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,0,2)  
    <<"\n\n";  
    fout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,1,2)  
    <<"\n\n";  
    fout<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,0)<<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,1)  
    <<" "<<CV_MAT_ELEM(*intrinsicR_matrixR,float,2,2)  
    <<"\n\n";  
  
  
f_right[0]=CV_MAT_ELEM(*intrinsicR_matrixR,float,0,0);  
f_right[1]=CV_MAT_ELEM(*intrinsicR_matrixR,float,1,1);  
  
  
    cout<<"畸变系数矩阵为：\n";  
    cout<<CV_MAT_ELEM(*distortionR_coeffsR,float,0,0)<<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,1,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,2,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,3,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,4,0)  
    <<"\n\n";  
  
  
fout<<"畸变系数矩阵为：\n";  
    fout<<CV_MAT_ELEM(*distortionR_coeffsR,float,0,0)<<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,1,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,2,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,3,0)  
    <<" "<<CV_MAT_ELEM(*distortionR_coeffsR,float,4,0)  
    <<"\n\n";  
  
  
    cvSave("intrinsicRs.xml",intrinsicR_matrixR);  
    cvSave("distortionR.xml",distortionR_coeffsR);  
  
  
    cout<<"摄像机矩阵、畸变系数向量已经分别存储在名为intrinsicRs.xml、distortionR.xml文档中\n\n";  
fout<<"摄像机矩阵、畸变系数向量已经分别存储在名为intrinsicRs.xml、distortionR.xml文档中\n\n";  
for(int ii = 0; ii < NImagesR; ++ii)  
{ float tranvR[3] = {0.0};  
float rotvR[3] = {0.0};  
  
  
for ( int i = 0; i < 3; i++)  
{  
tranvR[i] = ((float*)(translation_vectorsR->data.ptr+ii*translation_vectorsR->step))[i];  
rotvR[i] = ((float*)(rotation_vectorsR->data.ptr+rotation_vectorsR->step))[i];  
}  
  
cout << "第" << ii+1 << "幅图的外参数" << endl;  
printf("ROTATION VECTOR 旋转向量 : \n");  
printf("[ %6.4f %6.4f %6.4f ] \n", rotvR[0], rotvR[1], rotvR[2]);  
printf("TRANSLATION VECTOR 平移向量: \n");  
printf("[ %6.4f %6.4f %6.4f ] \n", tranvR[0], tranvR[1], tranvR[2]);  
printf("-----------------------------------------\n");  
  
  
fout << "第" << ii+1 << "幅图的外参数" << endl;  
fout<<"ROTATION VECTOR 旋转向量 : \n";  
fout<<"[ %6.4f %6.4f %6.4f ] \n"<<rotvR[0]<<"  "<< rotvR[1]<<"  "<< rotvR[2]<<endl;  
fout<<"TRANSLATION VECTOR 平移向量: \n";  
fout<<"[ %6.4f %6.4f %6.4f ] \n"<< tranvR[0]<<"  "<< tranvR[1]<<"  "<< tranvR[2]<<endl;  
fout<<"-----------------------------------------\n";  
  
  
}  
    CvMat * intrinsicR=(CvMat *)cvLoad("intrinsicRs.xml");  
    CvMat * distortionR=(CvMat *)cvLoad("distortionR.xml");  
  
  
    IplImage * mapxR=cvCreateImage(cvGetSize(showR_colieR),IPL_DEPTH_32F,1);  
    IplImage * mapyR=cvCreateImage(cvGetSize(showR_colieR),IPL_DEPTH_32F,1);  
  
  
    cvInitUndistortMap(intrinsicR,distortionR,mapxR,mapyR);  
  
  
    cvNamedWindow("原始图像右",1);  
    cvNamedWindow("非畸变图像右",1);  
  
  
   // cout<<"按‘E’键退出显示...\n\n";  
  
  
    /*while(showR_colieR) 
    {*/  
        IplImage * cloneR=cvCloneImage(showR_colieR);  
        cvShowImage("原始图像右",showR_colieR);  
        cvRemap(cloneR,showR_colieR,mapxR,mapyR);  
        cvReleaseImage(&cloneR);  
        cvShowImage("非畸变图像右",showR_colieR);  
cvWaitKey(500);  
        /*if(cvWaitKey(10)=='e') 
        { 
            break; 
        }*/  
  
  
        /*showR_colieR=cvQueryFrame(capture1); 
    }*/  
time_t start=0,end=0;  
int i,j,k;  
int max=-1;  
float hh[8];  
start=time(0);  
float para[8][9];  
float h[8];  
//CvCapture* cap1;  
//CvCapture* cap2;  
//Mat input11,input22;  
//cap1=cvCreateCameraCapture(0);  
//cap2=cvCreateCameraCapture(1);  
/*while(1) 
{ 
input11=cvQueryFrame(cap1); 
input22=cvQueryFrame(cap2); 
 
 
namedWindow("Camera_1",1); 
imshow("Camera_1",input11); 
namedWindow("Camera_2",1); 
imshow("Camera_2",input22); 
*/  
Mat input11,input22;  
//Mat input11=imread("123.jpg",1);  
//Mat input22=imread("124.jpg",1);  
//Mat input11(show_colie);  
//Mat input22(showR_colieR);  
Mat input1,input2;  
input11=imread("left_2.jpg",1);  
input22=imread("right_2.jpg",1);  
cvtColor(input11,input1,CV_RGB2GRAY,1);  
cvtColor(input22,input2,CV_RGB2GRAY,1);  
imwrite("left_2heibei.jpg",input1);  
imwrite("right_2heibai.jpg",input2);  
SiftFeatureDetector detector;  
vector<KeyPoint> keypoint1,keypoint2;  
detector.detect(input1,keypoint1);  
  
  
Mat output1;  
drawKeypoints(input1,keypoint1,output1);  
imshow("sift_result1.png",output1);  
imwrite("sift_result1.png",output1);  
  
  
Mat output2;  
SiftDescriptorExtractor extractor;  
Mat descriptor1,descriptor2;  
BruteForceMatcher<L2<float>> matcher;  
  
  
vector<DMatch> matches;  
Mat img_matches;  
detector.detect(input2,keypoint2);  
drawKeypoints(input2,keypoint2,output2);  
imshow("sift_result2.png",output2);  
imwrite("sift_result2.png",output2);  
  
  
extractor.compute(input1,keypoint1,descriptor1);  
extractor.compute(input2,keypoint2,descriptor2);  
  
  
matcher.match(descriptor1,descriptor2,matches);  
  
  
drawMatches(input1,keypoint1,input2,keypoint2,matches,img_matches);  
imshow("matches",img_matches);  
imwrite("matches.png",img_matches);  
//分配空间  
int pointcount=(int)matches.size();  
Mat point1(pointcount,2,CV_32F);  
Mat point2(pointcount,2,CV_32F);  
//把Keypoint转换为Mat  
Point2f point;  
for (i=0;i<pointcount;i++)  
{  
point=keypoint1[matches[i].queryIdx].pt;  
point1.at<float>(i,0)=point.x;  
point1.at<float>(i,1)=point.y;  
  
  
point=keypoint2[matches[i].trainIdx].pt;  
point2.at<float>(i,0)=point.x;  
point2.at<float>(i,1)=point.y;  
}  
//用RANSAC方法计算F  
Mat m_fundamental;  
vector<uchar> m_ransacstatus;  
  
  
m_fundamental=findFundamentalMat(point1,point2,m_ransacstatus,FM_RANSAC);  
/* 
float hhh[9]; 
for(i=0;i<9;i++) 
hhh[i]=0; 
for(i=0;i<3;i++) 
{ 
for (j=0;j<3;j++) 
{ 
hhh[i*3+j]=m_fundamental.ptr<float>(i)[j]; 
} 
} 
*/  
//计算外点个数  
int outlinercount=0;  
for(i=0;i<pointcount;i++)  
{  
if(m_ransacstatus[i]==0)//动态为0表示外点  
{  
outlinercount++;  
}  
}  
//计算内点  
vector<Point2f> m_leftinliner;  
vector<Point2f> m_rightinliner;  
vector<DMatch> m_inlinermatches;  
//上面三个变量用于保存内点和匹配关系  
int inlinercount=pointcount-outlinercount;  
m_inlinermatches.resize(inlinercount);  
m_leftinliner.resize(inlinercount);  
m_rightinliner.resize(inlinercount);  
  
  
inlinercount=0;  
for(i=0;i<pointcount;i++)  
{  
if(m_ransacstatus[i]!=0)  
{  
m_leftinliner[inlinercount].x=point1.at<float>(i,0);  
m_leftinliner[inlinercount].y=point1.at<float>(i,1);  
m_rightinliner[inlinercount].x=point2.at<float>(i,0);  
m_rightinliner[inlinercount].y=point2.at<float>(i,1);  
m_inlinermatches[inlinercount].queryIdx=inlinercount;  
m_inlinermatches[inlinercount].trainIdx=inlinercount;  
inlinercount++;  
}  
}  
//把内点转换为drawMatches可以使用的格式  
vector<KeyPoint> key1(inlinercount);  
vector<KeyPoint> key2(inlinercount);  
KeyPoint::convert(m_leftinliner,key1);  
KeyPoint::convert(m_rightinliner,key2);  
//显示计算F过后的内点匹配  
  
Mat outimage;  
  
drawMatches(input11,key1,input22,key2,m_inlinermatches,outimage);  
// namedWindow("Match features");  
imshow("Match feature",outimage);  
imwrite("提纯后的.jpg",outimage);  
vector<KeyPoint> ransac1,ransac2;  
vector<KeyPoint> left,right;  
vector<int> tichunyihou;  
int counter=0;  
int counterdidi=0;  
srand(unsigned(time(0)));  
int counterwo;  
int *countergege;  
countergege=new int[inlinercount];  
int * zuizhong=new int[inlinercount];  
for(i=0;i<inlinercount;i++)  
{  
zuizhong[i]=0;  
}  
left.clear();  
right.clear();  
  
  
while(counter<250)  
{  
counterwo=0;  
for(i=0;i<inlinercount;i++)  
{  
countergege[i]=0;  
}  
  
counterdidi++;  
ransac1.clear();  
ransac2.clear();  
  
  
int temp[4];  
for (i=0;i<4;i++)  
temp[i]=rand()%inlinercount;  
int gibal=0;  
  
  
for (i=0;i<3;i++)  
{  
for (j=i+1;j<4;j++)  
{  
if (temp[i]==temp[j])  
{  
gibal=1;  
break;  
}  
}  
}  
if(gibal==1)  
continue;  
int a,b;  
float tan1=0,tan2=0;  
for(i=0;i<3;i++)  
{  
j=i+1;  
a=(j-1+4)%4;  
b=(j+1+4)%4;  
tan1=(key1[m_inlinermatches[j].queryIdx].pt.y-key1[m_inlinermatches[a].queryIdx].pt.y)  
/(key1[m_inlinermatches[j].queryIdx].pt.x-key1[m_inlinermatches[a].queryIdx].pt.x);  
tan2=(key1[m_inlinermatches[j].queryIdx].pt.y-key1[m_inlinermatches[b].queryIdx].pt.y)  
/(key1[m_inlinermatches[j].queryIdx].pt.x-key1[m_inlinermatches[b].queryIdx].pt.x);  
  
  
if(abs(tan1/tan2-1)<=0.05)  
gibal=2;  
  
}  
if (gibal==2)  
continue;  
  
  
for (i=0;i<4;i++)  
{  
cout<<temp[i]<<"  ";  
fout<<temp[i]<<"  ";  
}  
cout<<endl;  
fout<<endl;  
if(gibal==1)  
continue;  
  
for(j=0;j<4;j++)  
{   
ransac1.push_back(key1[m_inlinermatches[temp[j]].queryIdx]);  
ransac2.push_back(key2[m_inlinermatches[temp[j]].trainIdx]);  
}  
if(ransac1.empty() || ransac2.empty())  
{  
cout<<"随机抽出四个特征点匹配对失败，请重试"<<endl;  
fout<<"随机抽出四个特征点匹配对失败，请重试"<<endl;  
continue;  
}  
//求透视矩阵  
  
for(i=0;i<8;i++)  
h[i]=0.0;  
for (i=0;i<=7;i++)  
{  
for (j=0;j<=8;j++)  
{  
para[i][j]=0;  
}  
}  
  
  
for (i=0;i<=2;i++)  
{  
para[i][2]=1;  
para[i+3][5]=1;  
para[i][0]=ransac1[i].pt.x;  
para[i][1]=ransac1[i].pt.y;  
para[i+3][3]=ransac1[i].pt.x;  
para[i+3][4]=ransac1[i].pt.y;  
para[i][6]=-ransac1[i].pt.x*ransac2[i].pt.x;  
para[i][7]=-ransac1[i].pt.y*ransac2[i].pt.x;  
para[i][8]=ransac2[i].pt.x;  
para[i+3][6]=-ransac1[i].pt.x*ransac2[i].pt.y;  
para[i+3][7]=-ransac1[i].pt.y*ransac2[i].pt.y;  
para[i+3][8]=ransac2[i].pt.y;  
}  
para[6][2]=1;  
para[7][5]=1;  
para[6][0]=ransac1[3].pt.x;  
para[6][1]=ransac1[3].pt.y;  
para[7][3]=ransac1[3].pt.x;  
para[7][4]=ransac1[3].pt.y;  
para[6][6]=-ransac1[3].pt.x*ransac2[3].pt.y;  
para[6][7]=-ransac1[3].pt.y*ransac2[3].pt.x;  
para[6][8]=ransac2[3].pt.x;  
para[7][6]=-ransac1[3].pt.x*ransac2[3].pt.y;  
para[7][7]=-ransac1[3].pt.y*ransac2[3].pt.y;  
para[7][8]=ransac2[3].pt.y;  
  
  
/////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////  
  
//建立完扩展矩阵，求出透视矩阵的参数：透视矩阵参数估值  
  
for (int g=0;g<8;g++)//初始化  
h[g]=0.0;  
for (i=0;i<8;i++)  
{  
float temper1=para[i][i];  
if (para[i][i]!=0)  
{  
for (int j=i+1;j<8;j++)  
{  
float temper2=para[j][i];  
if (para[j][i]!=0)  
{   
for (int k=i;k<9;k++)  
{  
para[j][k]=para[i][k]-para[j][k]*para[i][i]/temper2;  
}  
}  
else continue;  
  
}  
for (int r=i;r<9;r++)  
para[i][r]/=temper1;  
}  
else continue;  
}   
  
  
for (i=0;i<=7;i++)  
h[i]=para[i][8];  
for (j=7;j>=0;j--)  
{  
for (int k=0;k<7-j;k++)  
{  
h[j]-=h[7-k]*para[j][7-k];  
}  
}  
  
//计算出每个透视矩阵所满足的内点数  
float xx,yy;  
float juli;  
int ok=0;  
  
for(i=0;i<inlinercount;i++)  
{  
xx=0;yy=0;  
xx=h[0]*key1[m_inlinermatches[i].queryIdx].pt.x+h[1]*key1[m_inlinermatches[i].queryIdx].pt.y+h[2];  
yy=h[3]*key1[m_inlinermatches[i].queryIdx].pt.x+h[4]*key1[m_inlinermatches[i].queryIdx].pt.y+h[5];  
  
  
juli=sqrt(pow((xx-key2[m_inlinermatches[i].trainIdx].pt.x),2)+pow((yy-key2[m_inlinermatches[i].trainIdx].pt.y),2));  
if (juli<2)  
{  
  
countergege[counterwo]=i;  
counterwo++;  
ok++;  
}  
}  
  
  
if(ok>max)  
{  
max=ok;  
for(k=0;k<8;k++)  
hh[k]=h[k];  
for(i=0;i<inlinercount;i++)  
{  
zuizhong[i]=countergege[i];  
}  
}  
  
counter++;  
  
}  
for(i=0;i<8;i++)  
{  
cout<<hh[i]<<endl;  
fout<<hh[i]<<endl;  
}  
  
for(i=0;i<max;i++)  
{  
left.push_back(key1[m_inlinermatches[zuizhong[i]].queryIdx]);  
right.push_back(key2[m_inlinermatches[zuizhong[i]].trainIdx]);  
}  
cout<<"%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%"<<endl;  
vector<DMatch> wanquantichun;  
int haha=left.size();  
wanquantichun.resize(haha);  
  
for(i=0;i<haha;i++)  
{  
wanquantichun[i].queryIdx=i;  
wanquantichun[i].trainIdx=i;  
}  
Mat outimagehaha;  
float f=85;  
float T=2000;  
drawMatches(input11,left,input22,right,wanquantichun,outimagehaha);  
float xl=0,xr=0;  
float Z=430;  
for(i=0;i<haha;i++)  
{  
Z=0;  
if(key1[i].pt.x<428 && key1[i].pt.x>213 )  
{  
if(key2[i].pt.x<428 && key2[i].pt.x>213)  
{  
xl=abs(385-key1[i].pt.x);  
xr=abs(282-key2[i].pt.x);  
  
  
Z=abs(f*T/(xl+xr));  
cout<<"距离为=="<<Z<<endl;  
fout<<"距离为=="<<Z<<endl;  
}  
}  
}  
  
// namedWindow("Match features");  
/*for(i=0;i<haha;i++) 
{ 
 
if(left[i].pt.x<640 && left[i].pt.x>385 ) 
{ 
if(right[i].pt.x<282 && right[i].pt.x>0 ) 
{ 
xl=abs(385-key1[i].pt.x); 
xr=abs(282-key2[i].pt.x); 
 
 
//Z=abs(f*T/(xl-xr)); 
f=abs(Z*(xl+xr)/T); 
cout<<"距离为=="<<f<<endl; 
fout<<"距离为=="<<f<<endl; 
} 
} 
} 
*/  
imshow("完全提纯后  Match feature",outimagehaha);  
imwrite("完全提纯后的.jpg",outimagehaha);  
  
  
float uxleft,uxright;  
//uxleft=  
  
  
//图像配准  
int xx1=input11.cols;  
int yy1=input11.rows;  
int xx2=input22.cols;  
int yy2=input22.rows;  
int garo=0;  
int saero=0;  
int yiweiX=1000,yiweiYmin=1000,yiweiYmax=-1000;  
int xxx,yyy;  
for(i=0;i<yy1;i++)  
{  
for (j=0;j<xx1;j++)  
{  
garo=int(j*hh[0]+i*hh[1]+hh[2]);  
saero=int(j*hh[3]+i*hh[4]+hh[5]);  
if(yiweiX>garo)  
yiweiX=garo;  
if(yiweiYmin>saero)  
yiweiYmin=saero;  
if(yiweiYmax<saero)  
yiweiYmax=saero;  
}  
}  
  
if(yiweiX>0)  
yiweiX=0;  
  
cout<<yiweiYmin<<"      "<<yiweiX<<"     "<<yiweiYmax<<endl;  
fout<<yiweiYmin<<"      "<<yiweiX<<"     "<<yiweiYmax<<endl;  
  
  
if(yiweiYmax>=yy2)  
{  
//^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^1111111111111111111111111  
if(yiweiYmin<0)  
{  
cout<<"第一种情况"<<endl;  
fout<<"第一种情况"<<endl;  
Mat zuihoupeizhun(yiweiYmax-yiweiYmin,xx2-yiweiX,CV_8UC3,Scalar(0,0,0));  
for(i=0;i<yy1;i++)  
{  
for(j=0;j<xx1;j++)  
{  
garo=int(j*hh[0]+i*hh[1]+hh[2]);  
saero=int(j*hh[3]+i*hh[4]+hh[5]);  
xxx=garo-yiweiX;  
yyy=saero-yiweiYmin;  
if (yyy>=0 && xxx>=0&&xxx<(xx2-yiweiX)&&yyy<(yiweiYmax-yiweiYmin))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input11.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input11.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input11.at<Vec3b>(i,j)[2];  
}  
else   
continue;  
  
}   
}  
for (i=0;i<yy2;i++)  
{  
for (j=0;j<xx2;j++)  
{  
xxx=j-yiweiX;  
yyy=i-yiweiYmin;  
if (yyy>=0 && xxx>=0&&xxx<(xx2-yiweiX)&&yyy<(yiweiYmax-yiweiYmin))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input22.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input22.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input22.at<Vec3b>(i,j)[2];  
}  
else  
 continue;  
}  
}  
namedWindow("src");  
imshow("src",zuihoupeizhun);  
imwrite("src.png",zuihoupeizhun);  
}  
//^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^222222222222222  
else  
{  
cout<<"第二种情况"<<endl;  
fout<<"第二种情况"<<endl;  
Mat zuihoupeizhun(yiweiYmax,xx2-yiweiX,CV_8UC3,Scalar(0,0,0));  
for(i=0;i<yy1;i++)  
{  
for(j=0;j<xx1;j++)  
{  
garo=int(j*hh[0]+i*hh[1]+hh[2]);  
saero=int(j*hh[3]+i*hh[4]+hh[5]);  
xxx=garo-yiweiX;  
yyy=saero;  
if (yyy>=0 && xxx>=0 && yyy<yiweiYmax && xxx<(xx2-yiweiX))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input11.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input11.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input11.at<Vec3b>(i,j)[2];  
}  
else   
continue;  
  
}  
  
}  
for (i=0;i<yy2;i++)  
{  
for (j=0;j<xx2;j++)  
{  
xxx=j-yiweiX;  
yyy=i;  
if (yyy>=0 && xxx>=0 && yyy<yiweiYmax && xxx<(xx2-yiweiX))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input22.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input22.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input22.at<Vec3b>(i,j)[2];  
}  
else  
 continue;  
}  
}  
namedWindow("src");  
imshow("src",zuihoupeizhun);  
imwrite("src.png",zuihoupeizhun);  
}  
}  
else  
{  
//^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^3333333333333333333333333333  
if(yiweiYmin<0)  
{  
cout<<"第三种情况"<<endl;  
fout<<"第三种情况"<<endl;  
Mat zuihoupeizhun(yy2-yiweiYmin,xx2-yiweiX,CV_8UC3,Scalar(0,0,0));  
for(i=0;i<yy1;i++)  
{  
for(j=0;j<xx1;j++)  
{  
garo=int(j*hh[0]+i*hh[1]+hh[2]);  
saero=int(j*hh[3]+i*hh[4]+hh[5]);  
xxx=garo-yiweiX;  
yyy=saero-yiweiYmin;  
if (yyy>=0 && xxx>=0 && yyy<(yy2-yiweiYmin) && xxx<(xx2-yiweiX))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input11.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input11.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input11.at<Vec3b>(i,j)[2];  
}  
else   
continue;  
  
}   
}  
for (i=0;i<yy2;i++)  
{  
for (j=0;j<xx2;j++)  
{  
xxx=j-yiweiX;  
yyy=i-yiweiYmin;  
if (yyy>=0 && xxx>=0 && yyy<(yy2-yiweiYmin) && xxx<(xx2-yiweiX))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input22.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input22.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input22.at<Vec3b>(i,j)[2];  
}  
else  
 continue;  
}  
}  
namedWindow("src");  
imshow("src",zuihoupeizhun);  
imwrite("src.png",zuihoupeizhun);  
}  
else  
{  
//^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^444444444444444444  
cout<<"第四种情况"<<endl;  
fout<<"第四种情况"<<endl;  
Mat zuihoupeizhun(yy2,xx2-yiweiX,CV_8UC3,Scalar(0,0,0));  
for(i=0;i<yy1;i++)  
{  
for(j=0;j<xx1;j++)  
{  
garo=int(j*hh[0]+i*hh[1]+hh[2]);  
saero=int(j*hh[3]+i*hh[4]+hh[5]);  
xxx=garo-yiweiX;  
yyy=saero;  
if (yyy>=0 && xxx>=0 && yyy<yy2 && xxx<(xx2-yiweiX))  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input11.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input11.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input11.at<Vec3b>(i,j)[2];  
}  
else   
continue;  
  
}   
}  
for (i=0;i<yy2;i++)  
{  
for (j=0;j<xx2;j++)  
{  
xxx=j-yiweiX;  
yyy=i;  
if (xxx>=0 && yyy>=0)  
{  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[0]=input22.at<Vec3b>(i,j)[0];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[1]=input22.at<Vec3b>(i,j)[1];  
zuihoupeizhun.at<Vec3b>(yyy,xxx)[2]=input22.at<Vec3b>(i,j)[2];  
}  
else  
 continue;  
}  
}  
namedWindow("src");  
imshow("src",zuihoupeizhun);  
imwrite("src.jpg",zuihoupeizhun);  
}  
}  
  
end=time(0);  
cout<<"i love you"<<endl;  
fout<<"i love you"<<endl;  
  
  
  
  
cout<<"最后的max===="<<max<<endl;  
fout<<"最后的max===="<<max<<endl;  
  
end=time(0);  
cout<<endl<<"总跑程序的时间为："<<end-start<<"秒"<<endl;  
fout<<endl<<"总跑程序的时间为："<<end-start<<"秒"<<endl;  
cout<<endl;fout<<endl;  
cout<<"counter=="<<counter<<endl<<"counterdidi=="<<counterdidi<<endl;  
fout<<"counter=="<<counter<<endl<<"counterdidi=="<<counterdidi<<endl;  
waitKey(30000);  
}