OPENCV已知内参求外参

利用OPENCV，已知内参标定外参

#include <opencv2/calib3d.hpp>
#include <iostream>
#include <opencv2/opencv.hpp>
#include <fstream>
using namespace std;
using namespace cv;
int main() {//相机内参矩阵Mat cameraMatrix = Mat::eye(3, 3, CV_64F);cameraMatrix.at<double>(0, 0) = 1505.891829;cameraMatrix.at<double>(0, 1) = 0;cameraMatrix.at<double>(0, 2) = 973.395;cameraMatrix.at<double>(1, 1) = 1505.843003;cameraMatrix.at<double>(1, 2) = 597.616;cameraMatrix.at<double>(2, 2) = 1;//相机畸变系数Mat distCoeffs = Mat::zeros(5, 1, CV_64F);distCoeffs.at<double>(0, 0) = 0.0019057;distCoeffs.at<double>(1, 0) = 0;distCoeffs.at<double>(2, 0) = 0;distCoeffs.at<double>(3, 0) = 0;distCoeffs.at<double>(4, 0) = 0;//将控制点在世界坐标系的坐标压入容器vector<Point3f> objP;objP.clear();objP.push_back(Point3f(0, 4, 0));objP.push_back(Point3f(110, 65, 0));objP.push_back(Point3f(220, 4, 0));objP.push_back(Point3f(220, 176, 0));objP.push_back(Point3f(110, 115, 0));objP.push_back(Point3f(0, 176, 0));//6objP.push_back(Point3f(110, 0, 0));objP.push_back(Point3f(110, 180, 0));//8objP.push_back(Point3f(220, 50, 0));objP.push_back(Point3f(220, 130, 0));objP.push_back(Point3f(0, 130, 0));objP.push_back(Point3f(0, 50, 0));//12//将之前已经检测到的角点的坐标压入容器std::vector<Point2f> points;points.clear();points.push_back(Point2f(309, 43));points.push_back(Point2f(922, 401));points.push_back(Point2f(1595, 46));points.push_back(Point2f(1558, 1077));points.push_back(Point2f(915, 703));points.push_back(Point2f(301, 1016));//6points.push_back(Point2f(932, 13));points.push_back(Point2f(909, 1076));//8points.push_back(Point2f(1591, 320));points.push_back(Point2f(1576, 805));points.push_back(Point2f(296, 763));points.push_back(Point2f(300, 305));//12//因为坐标系的原因可能要手动镜像//for (int i = 0; i < points.size(); i++)//{//    /*int j = points[i].x;//    points[i].x = points[i].y;*///    points[i].y =  1200- points[i].y;//}//创建旋转矩阵和平移矩阵Mat rvecs = Mat::zeros(3, 1, CV_64FC1);Mat tvecs = Mat::zeros(3, 1, CV_64FC1);//求解pnp//solvePnP(objP, points, cameraMatrix, distCoeffs, rvecs, tvecs);solvePnPRansac(objP, points, cameraMatrix, distCoeffs, rvecs, tvecs);Mat rotM = Mat::eye(3, 3, CV_64F);Mat rotT = Mat::eye(3, 3, CV_64F);Rodrigues(rvecs, rotM);  //将旋转向量变换成旋转矩阵Rodrigues(tvecs, rotT);//计算相机旋转角double theta_x, theta_y, theta_z;double PI = 3.1415926;theta_x = atan2(rotM.at<double>(2, 1), rotM.at<double>(2, 2));theta_y = atan2(-rotM.at<double>(2, 0),sqrt(rotM.at<double>(2, 1) * rotM.at<double>(2, 1) + rotM.at<double>(2, 2) * rotM.at<double>(2, 2)));theta_z = atan2(rotM.at<double>(1, 0), rotM.at<double>(0, 0));theta_x = theta_x * (180 / PI);theta_y = theta_y * (180 / PI);theta_z = theta_z * (180 / PI);//计算深度Mat P;P = (rotM.t()) * tvecs;//输出cout << "角度" << endl;cout << theta_x << endl;cout << theta_y << endl;cout << theta_z << endl;cout << P << endl;return 0;
}

OPENCV已知内参求外参相关推荐

opencv已知相机投影及内参求外参_一步步用c++实现相机标定（张氏标定法）
enazoe/camera_calibration_cppgithub.com 前言最近在做相机标定方面的工作,虽然以前多次进行相机标定,但是多数时候是调用opencv的函数,过程相对简单.虽然对 ...
（九）相机内参、外参、反透视变换python opencv
背景知识任务需求:将相机上的一个点投影到真实世界平面上去. 原则上单目相机是不可以的,因为只记录了二维信息,真实世界是三维的,双目相机可以通过视差,或者单目+IMU组合,但是由于特征点在地面上的先验 ...
内参、外参、畸变参数三种参数与相机的标定方法与相机坐标系的理解
版权声明:本文为博主原创文章,未经博主允许不得转载. https://blog.csdn.net/yangdashi888/article/details/51356385 1. 相机参数是三种不同的 ...
计算机视觉（相机标定；内参；外参；畸变系数）
目录一.预备知识 1.坐标系变换过程(相机成像过程) (1)相机坐标系转换为图像坐标系(透视投影变换遵循的是针孔成像原理) (2)齐次坐标的引入原因:(为什么引入齐次坐标???) 2.内参与外参矩阵 ...
相机校正与相机内参、外参
简介: 对于相机进行校正,是为之后视觉测量奠定基础.本文在总结了相机校正中的基本概念.使用cv2的函数说明校正的过程. 关键词: 相机校正,内参,外参 #mermaid-svg-0bogxxocrI ...
使用ros标定相机的内参和外参
使用ros标定相机的内参和外参 1 安装依赖 2 运行标定节点 2.1 rocore 2.2 启动摄像头驱动程序 2.3 加载将要标定的图像主题 3 开始标定 3.1 打印一张棋盘格图片 3.2 具体 ...
已知坐标求方位角_【干货】RTK视频实操 | 求转换参数详解和七点注意事项！
RTK视频教学,『求转换参数』,技术员结合测量一线实操经验,运用工程之星5.0软件为大家讲解如何求转换参数,以及求转换参数需要注意哪些事项. 求转换参数操作通常情况下,科力达工程之星直接输出的坐标为 ...
Realsense d435i内参、外参标定
使用工具code_utils.imu_utils.kalibr对Realsense d435i 内参.外参标定.本文介绍上述工具的安装,及标定方法. 一.code_utils安装 1.建立工作空间 m ...
ros --- 双目相机内参与外参标定
ros --- 双目相机内参与外参标定小觅相机直接获取参数手动重新标定 1. 双目相机内外参标定生成标定板录制 stereo_calibra.bag 标定标定结果标定验证 2. 双目 + ...

OPENCV已知内参求外参

OPENCV已知内参求外参相关推荐

最新文章

热门文章