相机标定opencv实现

标定图片（本科时拍摄

 

 

 

 

 

 

 

标定代码

#include <cv.h>
#include <highgui.h>
#include <stdio.h>
#include <stdlib.h>
#include <iostream>
#include <opencv2/opencv.hpp>

using namespace std;
using namespace cv;


class CCalibration
{
public:
    CCalibration(CvSize _board_sz, double _board_dt, int _n_boards = 15);
    ~CCalibration();

public:
    bool doCalibrate(const CvMat* const image_points, const CvMat* const object_points,const CvMat* const point_counts, CvSize size);
    bool calibrateFromCamera();
    bool calibrateFromFile();
    void display();

protected:

private:
    CvSize board_sz; //标定板信息
    int n_boards;    //视场总数
    double board_dt; //相邻视场间的获取时间间隔

private:

    CvMat* intrinsic_matrix;//内参数矩阵
    CvMat* distortion_coeffs;//畸变矩阵

};



CCalibration::CCalibration(CvSize _board_sz, double _board_dt, int _n_boards)
{
    //标定板的信息
    board_sz = _board_sz;
    board_dt = _board_dt;
    n_boards = _n_boards;

    //为标定参数分配内存
    intrinsic_matrix  = cvCreateMat(3,3,CV_32FC1);
    distortion_coeffs = cvCreateMat(4,1,CV_32FC1);
}

CCalibration::~CCalibration()
{
    cvReleaseMat(&intrinsic_matrix);
    cvReleaseMat(&distortion_coeffs);
}

/* 
* 函数名称：calibrateFromCamera
* 函数功能：直接从相机实时获取标定板图像，用于标定
* 函数入口：  
* 输入参数：五
* 输出参数：无
* 返 回 值：是否标定成功，true表示成功，false表示失败
* 其它说明：  
*/  
bool CCalibration::calibrateFromCamera()
{
    cvNamedWindow("Calibration",CV_WINDOW_AUTOSIZE);
    cvNamedWindow("Live",CV_WINDOW_AUTOSIZE);

    CvCapture* capture = cvCreateCameraCapture( 0 );//将要标定的摄像头
    assert( capture );

    int board_n = board_sz.width * board_sz.height;//角点总数
    CvMat* image_points      = cvCreateMat(n_boards*board_n,2,CV_32FC1);// cvMat* cvCreateMat ( int rows, int cols, int type )
    CvMat* object_points     = cvCreateMat(n_boards*board_n,3,CV_32FC1);//cvCreateMat预定义类型的结构如下：CV_<bit_depth> (S|U|F)C<number_of_channels>
    CvMat* point_counts      = cvCreateMat(n_boards,1,CV_32SC1);//cvCreateMat矩阵的元素可以是32位浮点型数据(CV_32FC1)，或者是无符号的8位三元组的整型数据(CV_8UC3)

    CvPoint2D32f* corners = new CvPoint2D32f[ board_n ];

    IplImage *image = cvQueryFrame( capture );
    //imgSize = cvGetSize(image);
    IplImage *gray_image = cvCreateImage(cvGetSize(image),8,1);//subpixel   创建单通道灰度图像

    int corner_count;
    int successes = 0;//图像系列index
    int step, frame = 0;

    //忽略开始前2s时间的图片
    for (int i = 0; i < 33*2; i++)
    {
        image = cvQueryFrame(capture);
        cvShowImage("Live",image);
        cvWaitKey(30);
    }
    //获取足够多视场图片用于标定
    while (successes < n_boards)
    {
        image = cvQueryFrame(capture);
        cvShowImage("Live", image);
        cvWaitKey(33);//一帧的时间间隔

        //每隔board_dt秒取一张图像
        if ( (frame++ % ((int)(33 * board_dt)) ) == 0 )
        {
            //Find chessboard corners:
            int found = cvFindChessboardCorners(image, board_sz, corners, &corner_count,
                CV_CALIB_CB_ADAPTIVE_THRESH | CV_CALIB_CB_FILTER_QUADS);
            if(found == 0)  continue;//未正确找到角点，继续下一次

            //Get Subpixel accuracy on those corners
            cvCvtColor(image, gray_image, CV_BGR2GRAY);                //转换为灰度图像
            cvFindCornerSubPix(gray_image, corners, corner_count,      //cvFindChessboardCorners找到的角点仅仅是近似值，必须调用此函数达到亚像素精度，如果第一次定位...
                cvSize(11,11),cvSize(-1,-1), cvTermCriteria(    //角点时忽略调用此函数，那么会导致标定的实际错误
                CV_TERMCRIT_EPS+CV_TERMCRIT_ITER, 30, 0.1 ));

            // 如果该视场获得了好的结果，保存它
            // If we got a good board, add it to our data
            if (corner_count == board_n)
            {
                step = successes*board_n;
                for( int i=step, j=0; j<board_n; ++i,++j ) 
                {
                    CV_MAT_ELEM(*image_points, float,i,0) = corners[j].x;  // CV_MAT_ELEM 用来访问矩阵每个元素的宏，这个宏只对单通道矩阵有效，多通道会报错...
                    CV_MAT_ELEM(*image_points, float,i,1) = corners[j].y;  //CV_MAT_ELEM( matrix, elemtype, row, col )
                    CV_MAT_ELEM(*object_points,float,i,0) = j/board_sz.width;     //matrix：要访问的矩阵,elemtype：矩阵元素的类型,row：所要访问元素的行数,col：所要访问元素的列数
                    CV_MAT_ELEM(*object_points,float,i,1) = j%board_sz.width;
                    CV_MAT_ELEM(*object_points,float,i,2) = 0.0f;
                }
                CV_MAT_ELEM(*point_counts, int,successes,0) = board_n;    
                successes++;
            }

            //Draw corners
            cvDrawChessboardCorners(image, board_sz, corners, corner_count, found);//found为cvFindChessboardCorners的返回值
            char text[10];
            sprintf(text,"%d/%d", successes,n_boards);
            CvFont font = cvFont(2,2);
            cvPutText(image,text,cvPoint(40,40),&font,cvScalar(0,0,255));
            cvShowImage( "Calibration", image );
        }
    }

    //获取了足够多视场，结束获取
    cvDestroyWindow("Calibration");
    cvDestroyWindow("Live");
    //计算
    doCalibrate(image_points, object_points, point_counts, cvGetSize(image));

    //结束
    delete []corners;
    cvReleaseMat(&image_points);
    cvReleaseMat(&object_points);
    cvReleaseMat(&point_counts);
    cvReleaseImage(&gray_image);
    cvReleaseCapture(&capture);

    return true;
}/*  calibrateFromCamera()   */

/* 
* 函数名称：calibrateFromCamera
* 函数功能：根据已获取的图像文件（.bmp格式），标定相机
* 函数入口：  
* 输入参数：无
* 输出参数：无
* 返 回 值：是否标定成功，true表示成功，false表示失败
* 其它说明： 只接受.bmp格式的图片，且图片尺寸要相同，若要标定其他格式图片，请将本函数内的.bmp替换成.jpg
*            文件统一命名格式为 calib_N.bmp,其中N必须从0开始
*/  
bool CCalibration::calibrateFromFile()
{
    cvNamedWindow("Calibration", CV_WINDOW_AUTOSIZE);
    //cvNamedWindow("FileImage", CV_WINDOW_AUTOSIZE);
    int board_n = board_sz.width * board_sz.height;//每张图角点总数
    CvMat* image_points      = cvCreateMat(n_boards*board_n,2,CV_32FC1);// cvMat* cvCreateMat ( int rows, int cols, int type )
    CvMat* object_points     = cvCreateMat(n_boards*board_n,3,CV_32FC1);//cvCreateMat预定义类型的结构如下：CV_<bit_depth> (S|U|F)C<number_of_channels>
    CvMat* point_counts      = cvCreateMat(n_boards,1,CV_32SC1);//cvCreateMat矩阵的元素可以是32位浮点型数据(CV_32FC1)，或者是无符号的8位三元组的整型数据(CV_8UC3)

    CvPoint2D32f* corners = new CvPoint2D32f[ board_n ];

    char imgName[20] = "1.jpg";//加载图
    IplImage *image = cvLoadImage(imgName,1);
    IplImage *gray_image = cvCreateImage(cvGetSize(image),8,1);//subpixel   创建单通道灰度图像

    int corner_count;
    int successes = 0, index = 1;//图像系列index
    int step;

    //获取足够多视场图片用于标定
    while (successes < n_boards)
    {
        sprintf(imgName, "%d.jpg",index++);        //一次加载n_boards张图片
        image = cvLoadImage(imgName,1);
        if ( !image ) break; //无此图片，则停止
        cvWaitKey(1000*board_dt);//一帧的时间间隔

        //Find chessboard corners:
        int found = cvFindChessboardCorners(image, board_sz, corners, &corner_count,
            CV_CALIB_CB_ADAPTIVE_THRESH | CV_CALIB_CB_FILTER_QUADS);
        if(found == 0)  continue;//未正确找到角点，继续下一次

        //Get Subpixel accuracy on those corners
        cvCvtColor(image, gray_image, CV_BGR2GRAY);                //转换为灰度图像
        cvFindCornerSubPix(gray_image, corners, corner_count,      //cvFindChessboardCorners找到的角点仅仅是近似值，必须调用此函数达到亚像素精度，如果第一次定位...
            cvSize(11,11),cvSize(-1,-1), cvTermCriteria(    //角点时忽略调用此函数，那么会导致标定的实际错误
            CV_TERMCRIT_EPS+CV_TERMCRIT_ITER, 30, 0.1 ));

        // 如果该视场获得了好的结果，保存它
        // If we got a good board, add it to our data
        if (corner_count == board_n)
        {
            step = successes*board_n;
            for( int i=step, j=0; j<board_n; ++i,++j ) 
            {
                CV_MAT_ELEM(*image_points, float,i,0) = corners[j].x;  // CV_MAT_ELEM 用来访问矩阵每个元素的宏，这个宏只对单通道矩阵有效，多通道会报错...
                CV_MAT_ELEM(*image_points, float,i,1) = corners[j].y;  //CV_MAT_ELEM( matrix, elemtype, row, col )
                CV_MAT_ELEM(*object_points,float,i,0) = j/board_sz.width;     //matrix：要访问的矩阵,elemtype：矩阵元素的类型,row：所要访问元素的行数,col：所要访问元素的列数
                CV_MAT_ELEM(*object_points,float,i,1) = j%board_sz.width;
                CV_MAT_ELEM(*object_points,float,i,2) = 0.0f;
            }
            CV_MAT_ELEM(*point_counts, int,successes,0) = board_n;    
            successes++;
        }

        //Draw corners
        cvDrawChessboardCorners(image, board_sz, corners, corner_count, found);//found为cvFindChessboardCorners的返回值
        char text[10];
        sprintf(text,"%d/%d", successes,n_boards);
        CvFont font = cvFont(2,2);
        cvPutText(image,text,cvPoint(40,40),&font,cvScalar(0,0,255));
        cvShowImage( "Calibration", image );
    }

    //获取了足够多视场，结束获取
    cvDestroyWindow("Calibration");
    //cvDestroyWindow("FileImage");
    doCalibrate(image_points, object_points, point_counts, cvGetSize(image));

    delete []corners;
    cvReleaseMat(&image_points);
    cvReleaseMat(&object_points);
    cvReleaseMat(&point_counts);
    cvReleaseImage(&image);
    cvReleaseImage(&gray_image);
    return true;
} /* calibrateFromFile() */

/* 
* 函数名称：doCalibrate 
* 函数功能：计算相机内参数和畸变参数 
* 函数入口：
* 输入参数：存储图像角点坐标（成像仪坐标）信息的矩阵指针image_points，存储有标定板角点坐标（世界坐标）信息的矩阵指针object_points
*            存储有各图像寻找到的角点个数信息的矩阵指针point_counts，图像尺寸size
* 输出参数：无 
* 返 回 值： 是否成功，true成功，false失败
* 其它说明： 标定结果同时存储到当前目录Intrinsics.xml，Distortion.xml文件中
*/  
bool CCalibration::doCalibrate(const CvMat* const image_points, const CvMat* const object_points,const CvMat* const point_counts, CvSize size)
{
    //****************************开始标定*************************//

    // 初始化内参数矩阵的fx和fy为1.0f
    CV_MAT_ELEM( *intrinsic_matrix, float, 0, 0 ) = 1.0f;
    CV_MAT_ELEM( *intrinsic_matrix, float, 1, 1 ) = 1.0f;

    //**************计算标定参数*************//
    //CALIBRATE THE CAMERA!
    cvCalibrateCamera2( object_points, image_points, point_counts,  size,
        intrinsic_matrix, distortion_coeffs,
        NULL, NULL,0  //CV_CALIB_FIX_ASPECT_RATIO
        );

    //SAVE THE INTRINSICS AND DISTORTIONS
    cvSave("Intrinsics.xml",intrinsic_matrix);//保存摄像头内参数
    cvSave("Distortion.xml",distortion_coeffs);//保存摄像头外参数

    return true;
}/* doCalibrate() */


/* 
* 函数名称：display 
* 函数功能：根据标定参数，显示修正后的视频图像 
* 函数入口：
* 输入参数：无
* 输出参数：无 
* 返 回 值： 
* 其它说明：  
*/  
void CCalibration::display()
{
    cvNamedWindow("Undistort", CV_WINDOW_AUTOSIZE);//显示修正后图像

    CvCapture *capture = cvCreateCameraCapture(0);
    IplImage *frame = cvQueryFrame(capture);
    IplImage *imgUndistort = cvCreateImage(cvGetSize(frame),frame->depth,frame->nChannels);

    // EXAMPLE OF LOADING THESE MATRICES BACK IN:
    CvMat *intrinsic = (CvMat*)cvLoad("Intrinsics.xml");//加载摄像头内参数
    CvMat *distortion = (CvMat*)cvLoad("Distortion.xml");//加载摄像头外参数

    // Build the undistort map which we will use for all subsequent frames.
    IplImage* mapx = cvCreateImage( cvGetSize(frame), IPL_DEPTH_32F, 1 );
    IplImage* mapy = cvCreateImage( cvGetSize(frame), IPL_DEPTH_32F, 1 );

    //计算畸变映射 即根据摄像头内、外参数，计算出如果没有这些畸变的话，摄像头获得的理想图像
    cvInitUndistortMap(intrinsic, distortion, mapx, mapy);

    while(cvWaitKey(33) != 27) //ESC
    {
        frame = cvQueryFrame(capture);
        cvRemap( frame, imgUndistort, mapx, mapy);
        cvShowImage("Undistort", imgUndistort);
    }

    cvReleaseCapture(&capture);
    cvReleaseImage(&imgUndistort);
    cvDestroyWindow("Undistort");

}/* display()  */

void main()
{
    //CCalibration calib(cvSize(7,8),1,10);
    CCalibration calib(cvSize(6,8),1,10);
    //从相机中获取图像标定
    calib.calibrateFromCamera();

    //从已有图像中标定
    //calib.calibrateFromFile();

    //运用标定结果显示修正后图像
    //calib.display();
    //system("pause");
}

相机参数

F:学科、技能、编程【】课程大四【】机器视觉及应用课程设计项目机器视觉项目设计（汤--组）相机参数

 

标定载物台的比例系数后即可测量零件尺寸