项目实战：Qt+Opencv相机标定工具v1.3.0（支持打开摄像头、视频文件和网络地址，支持标定过程查看、删除和动态评价误差率，支持追加标定等等）

若该文为原创文章，转载请注明出处
本文章博客地址：https://hpzwl.blog.csdn.net/article/details/141334834

长沙红胖子Qt（长沙创微智科）博文大全：开发技术集合（包含Qt实用技术、树莓派、三维、OpenCV、OpenGL、ffmpeg、OSG、单片机、软硬结合等等）持续更新中…

Qt开发专栏：项目实战（点击传送门）

需求

  1.打开摄像头，可设置帧率、分辨率（可设置）；
  2.可打开usb、rtsp和本地文件（直接输入地址自动判断）；
  3.opencv摄像头操作子线程处理；
  4.支持设置棋盘格的行列角点数；
  5.支持标定过程可控制；
  6.采集标定图、可对标定图进行查看、删除；
  7.可对已有的标定图查看评价像素误差率；
  8.标定完成后，可以追加标定，继续开始基于原来的标定采集图继续标定；
  9.支持定制配置文件的导出和导出（测试运行包不对外开放该功能）；

相关博客

  《OpenCV开发笔记（〇）：使用mingw530_32编译openCV3.4.1源码，搭建Qt5.9.3的openCV开发环境》
  《OpenCV开发笔记（三）：OpenCV图像的概念和基本操作》
  《OpenCV开发笔记（四）：OpenCV图片和视频数据的读取与存储》
  《OpenCV开发笔记（五）：OpenCV读取与操作摄像头》
  《OpenCV开发笔记（六）：OpenCV基础数据结构、颜色转换函数和颜色空间》
  《OpenCV开发笔记（七十六）：相机标定（一）：识别棋盘并绘制角点》
  《OpenCV开发笔记（七十七）：相机标定（二）：通过棋盘标定计算相机内参矩阵矫正畸变摄像头图像》

Demo：calibrateTool_v1.3.0 windows运行包

  广角摄像头标定过程
  

  

  鱼眼摄像头标定过程
  

  

  动态标定过程：查看、删除和评价
  

  

  CSDN粉丝0积分下载：https://download.csdn.net/download/qq21497936/89652658
  QQ群：博客首页扫码进入QQ技术群，点击“文件”搜索“calibrateTool”，群内与博文同步更新）

模块化部署

  

关键源码

CalibrateManager.h

#ifndef CALIBRATEMANAGER_H
#define CALIBRATEMANAGER_H// opencv
#include "opencv/highgui.h"
#include "opencv/cxcore.h"
#include "opencv2/core/core.hpp"
#include "opencv2/highgui/highgui.hpp"
#include "opencv2/opencv.hpp"
#include "opencv2/xphoto.hpp"
#include "opencv2/dnn/dnn.hpp"
// opencv_contrib
#include <opencv2/xphoto.hpp>
#include <opencv2/ximgproc.hpp>
#include <opencv2/calib3d.hpp>
#include <opencv2/features2d.hpp>
#include <opencv2/xfeatures2d.hpp>
#include <opencv2/xfeatures2d/nonfree.hpp>#include "cvui.h"
#include "calibrateCommon.h"
#include <QImage>
#include <QTimer>class CalibrateManager: public QObject
{Q_OBJECT
public:explicit CalibrateManager(QObject *parent = 0);~CalibrateManager();public slots:void testOpencvEnv();                       // 测试环境public:double  getBrightness()     const;          // 亮度double  getContrast()       const;          // 对比度double  getSaturation()     const;          // 饱和度double  getHue()            const;          // 色调double  getGain()           const;          // 增益double  getExposure()       const;          // 曝光度bool    getShowProperty()   const;          // 显示属性int     getCalibrateRegionX() const;        // 区域xint     getCalibrateRegionY() const;        // 区域yint     getCalibrateRegionWidth() const;    // 区域widthint     getCalibrateRegionHeight() const;   // 区域heightint     getChessboardColCornerCount() const;// 棋盘行角点数量int     getChessboardRowCornerCount() const;// 棋盘列角点数量QString getSerialize() const;               // 获取序列化参数public:void setBrightness  (double value);         // 亮度void setContrast    (double value);         // 对比度void setSaturation  (double value);         // 饱和度void setHue         (double value);         // 色调void setGain        (double value);         // 增益void setExposure    (double value);         // 曝光度void setShowProperty(bool value);           // 显示属性void setCalibrateRegionX(int x);            // 区域xvoid setCalibrateRegionY(int y);            // 区域yvoid setCalibrateRegionWidth(int width);    // 区域widthvoid setCalibrateRegionHeight(int height);  // 区域heightvoid setChessboardColCornerCount(int count);// 棋盘行角点数量void setChessboardRowCornerCount(int count);// 棋盘列角点数量bool setSerialize(QString str);             // 获取序列化参数signals:void signal_opened(bool result);            // 打开摄像头信号void signal_closed();                       // 关闭摄像头信号void signal_captureOneFrame(cv::Mat mat);   // 接收图像后抛出信号void signal_captureOneFrame(QImage image);  // 接收图像后抛出信号void signal_captureOneResultFrame(cv::Mat mat);   // 接收图像后抛出信号void signal_captureOneResultFrame(QImage image);  // 接收图像后抛出信号void signal_startedCalibrate(bool result);  // 开始标定结果void signal_regionChanged(int x, int y, int width, int height);void signal_fpsChanged(int fps);            // 帧率void signal_stopedCalibrate();              // 结束标定结果（这是强制中断，不是标定完成）void signal_finishedCalibrate();            // 标定完成void signal_cameraInfo(CameraInfo cameraInfo);  // 更新截图相机信息public slots:void slot_startCapture(int usb, int width = 0, int height = 0, int fps = 0);// 打开摄像头, 0...void slot_startCapture(QString url, int width = 0, int height = 0, int fps = 0);// 打开摄像头, 网络摄像头地址void slot_stopCapture();                    // 当正在采集中时(>>时)，关闭摄像头会导致程序崩溃，所以采集与停止放一个线程中(消息循环)void slot_startCalibrate();                 // 开始标定void slot_addCalibrate();                   // 继续标定void slot_snapshot();                       // 快照void slot_deleteSnapshot(int index);        // 删除快照void slot_stopCalibrate();                  // 停止标定void slot_finishCalibrate();                // 完成标定public slots:void slot_start();                       // 开启线程void slot_stop();                        // 关闭线程protected slots:void slot_captrueFrame();           // 消息循环获取图像protected:void initControl();void updateCalibrateResult();       // 更新标定结果void calculateCalibrateErrors();    // 计算误差protected:bool findChessboard(int rowCornerCount, int colCornerCount, cv::Mat &mat, std::vector<cv::Point2f> &vectorPoint2fCorners);public:static QImage mat2Image(cv::Mat mat);      // cv::Mat 转 QImageprivate:bool _running;                      // 线程是否运行private:cv::VideoCapture *_pVideoCapture;   // 摄像头实例bool _showProperty;                 // 是否显示属性参数double _brightness;                 // 亮度double _contrast;                   // 对比度double _saturation;                 // 饱和度double _hue;                        // 色调double _gain;                       // 增益double _exposure;                   // 曝光度int _width;                         // 宽度int _height;                        // 高度int _fps;                           // 帧率bool _opened;                       // 摄像头是否打开bool _calibratingBefore;            // 标定前一个变化状态bool _calibrating;                  // 正在标定bool _calibratFinished;             // 校准完了（当前最近一个已经校准）int _calibrateRegionX;              // 标定region区域像素起始x坐标int _calibrateRegionY;              // 标定region区域像素起始y坐标int _calibrateRegionWidth;          // 标定region区域像素宽度int _calibrateRegionHeight;         // 标定region区域像素高度cv::Mat _mat;                       // 缓存一帧cv::Mat _resultMat;                 // 结果int _chessboardColCornerCount;      // 一列多少个角点int _chessboardRowCornerCount;      // 一行多少个角点std::vector<std::vector<cv::Point3f>> _vectorObjectPoint;    // 缓存点std::vector<std::vector<cv::Point2f>> _vectorImagePoint;bool _snapshot;                     // 拍照private:                                // 计算内参和畸变系数cv::Mat _cameraMatrix;              // 相机矩阵（接收输出）cv::Mat _distCoeffs;                // 畸变系数（接收输出）std::vector<cv::Mat> _rotate;       // 旋转量（接收输出）std::vector<cv::Mat> _translate;    // 偏移量（接收输出）private:CameraInfo _cameraInfo;
};#endif // CALIBRATEMANAGER_H

CalibrateManager.cpp

...
void CalibrateManager::slot_captrueFrame()
{if(!_running){return;}if(_pVideoCapture->isOpened()){*_pVideoCapture >> _mat;if(_showProperty){cv::putText(_mat, QString("brightness: %1").arg(_brightness).toStdString(),cvPoint(0, 30), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("  contrast: %1").arg(_contrast  ).toStdString(),cvPoint(0, 60), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("saturation: %1").arg(_saturation).toStdString(),cvPoint(0, 90), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("       hue: %1").arg(_hue       ).toStdString(),cvPoint(0, 120), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("      gain: %1").arg(_gain      ).toStdString(),cvPoint(0, 150), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("  exposure: %1").arg(_exposure  ).toStdString(),cvPoint(0, 180), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));cv::putText(_mat, QString("press ESC out").toStdString(),cvPoint(0, 210), cv::FONT_HERSHEY_COMPLEX, 1.0, cv::Scalar(255));}// 第一次进入标定if(!_calibratingBefore && _calibrating){_calibrateRegionX = 0;_calibrateRegionY = 0;_calibrateRegionWidth = _width;_calibrateRegionHeight = _height;_calibratingBefore = true;emit signal_regionChanged(_calibrateRegionX, _calibrateRegionY, _calibrateRegionWidth, _calibrateRegionHeight);QImage srcImage = mat2Image(_mat);emit signal_captureOneResultFrame(srcImage);}else if(_calibrating){QImage srcImage = mat2Image(_mat);// 获取std::vector<cv::Point2f> imagePoints;if(findChessboard(_chessboardRowCornerCount,_chessboardColCornerCount,_mat,imagePoints)){// 这是拍照截图if(_snapshot){// 三维世界坐标系std::vector<cv::Point3f> objectPoints;for(int i = 0; i < _chessboardRowCornerCount; i++){for(int j = 0; j < _chessboardColCornerCount; j++){objectPoints.push_back(cv::Point3f(j, i, 0));}}// 图像识别出来的角点（一张图一组）_vectorObjectPoint.push_back(objectPoints);_vectorImagePoint.push_back(imagePoints);_snapshot = false;{SnapShot snapShot;snapShot.dateTime = QDateTime::currentDateTime().toString("yyyy-MM-dd hh:mm:ss:zzz");snapShot.srcImage = srcImage;snapShot.drawChessboardImage = mat2Image(_mat);snapShot.imagePoints = imagePoints;snapShot.objectPoints = objectPoints;_cameraInfo.listSnapShot.append(snapShot);// 更新标定结果updateCalibrateResult();// 计算误差率calculateCalibrateErrors();// 抛出更新emit signal_cameraInfo(_cameraInfo);}}}
//            if(_cameraInfo.listSnapShot.size() == 0)
//            {
//                QImage srcImage = mat2Image(_mat);
//                emit signal_captureOneResultFrame(srcImage);
//            }else{
//                cv::undistort(_mat, _resultMat, _cameraMatrix, _distCoeffs);
//                QImage image = mat2Image(_resultMat);
//                emit signal_captureOneResultFrame(image);
//            }}else if(_calibratFinished){
//            if(_cameraInfo.listSnapShot.size() == 0)
//            {
//                QImage srcImage = mat2Image(_mat);
//                emit signal_captureOneResultFrame(srcImage);
//            }else{
//                cv::undistort(_mat, _resultMat, _cameraMatrix, _distCoeffs);
//                QImage image = mat2Image(_resultMat);
//                emit signal_captureOneResultFrame(image);
//            }}// 抛出原图QImage image = mat2Image(_mat);emit signal_captureOneFrame(image);// 抛出校正图if(_cameraMatrix.empty()){emit signal_captureOneResultFrame(image);}else{LOG;cv::undistort(_mat, _resultMat, _cameraMatrix, _distCoeffs);QImage dstImage = mat2Image(_resultMat);emit signal_captureOneResultFrame(dstImage);}QTimer::singleShot(5, this, SLOT(slot_captrueFrame()));}
}
...

入坑

  算法的研究优化过程中，受到摄像头光学、标定板、标定板所占视口大小，图像处理过程原本的流程优化、标定过程中动态的处理等多方面因素，坑多暂时未记录。

本文章博客地址：https://hpzwl.blog.csdn.net/article/details/141334834