C++如何将欧拉角转为旋转矩阵

eigen 旋转矩阵　按照RxRyRz或者RzRyRx转欧拉角

欧拉角求旋转矩阵c++

在手眼标定中，现在有一组数据为[-0.3969707, -0.460018, 0.3899877, 90.2261, -168.2015, 89.7748]，分别代表机械手的位置XYZ和旋转角RXRYRZ，旋转方向为xyz，如何将它转为旋转矩阵和平移向量，结合opencv和eigen请用C++实现

2. **旋转矩阵**：将欧拉角（单位：度）按 XYZ 顺序转换为旋转矩阵 $$ R = R_z(\gamma) \cdot R_y(\beta) \cdot R_x(\alpha) $$ 其中： $$ R_x(\alpha) = \begin{bmatrix} 1 & 0 & 0 \\ 0 & \cos\alpha & -\sin\...

Qt中如何将旋转角度转为欧拉角

QMatrix4x4类用于将四元数转换为旋转矩阵，rotation()方法用于获取旋转矩阵中的旋转部分，toEulerAngles()方法用于将旋转矩阵中的旋转部分转换为欧拉角。最终，euler就是对应的欧拉角值。需要注意的是，欧拉角的...

c++ opengl四元数旋转

而相较于旋转矩阵，则更加节省存储空间并易于插值。 #### 使用 GLM 库创建和应用四元数 GLM 是一个广泛使用的开源数学库，专为基于 OpenGL 的应用程序设计。通过 GLM 可以轻松实现四元数的定义、初始化以及转换到...

#include <map> #include <vector> #include <Eigen/Eigen> #include <opencv2/opencv.hpp> #include <cv_bridge/cv_bridge.h> #include <iostream> #include <fstream> // 棋盘格检测器类，用于检测棋盘格并计算机器人/相机坐标系转换 class ChessboardDetector_cls { private: cv::Mat rvec; // 旋转向量（世界坐标系到相机坐标系的旋转） cv::Mat tvec; // 平移向量（世界坐标系原点到相机坐标系原点的平移） // 相机内参矩阵 [fx, 0, cx; 0, fy, cy; 0, 0, 1] cv::Mat cameraMatrix; // 畸变系数 [k1, k2, p1, p2, k3] cv::Mat distCoeffs; // 棋盘格方块物理尺寸（单位：米，需与实际应用场景一致） float squareSize; int boardWidth = 7; // 棋盘格宽度（单位：内角点） int boardHeight = 7; // 棋盘格高度（单位：内角点） bool cvVerPlus45 = false; // OpenCV版本判断标志 public: // 构造函数，初始化相机参数和棋盘格尺寸 ChessboardDetector_cls(float square_size = 12.5,int nDevType=1) // 默认1cm方块,1=r1d,0=r8c { if(nDevType == 0){ // ✅ 相机内参矩阵 cameraMatrix: // [[1.04969613e+03 0.00000000e+00 5.73570763e+02] // [0.00000000e+00 1.05086275e+03 4.05877726e+02] // [0.00000000e+00 0.00000000e+00 1.00000000e+00]] // ✅ 畸变系数 distCoeffs: // [[-1.85926395e-03 1.61417431e-03 1.14303737e-03 7.58650886e-05 // -1.17211371e-03]] this->cameraMatrix = (cv::Mat_<double>(3, 3) << 593.52455247, 0, 359.99576897, 0, 591.1469869, 231.48422218, 0, 0, 1); this->distCoeffs = (cv::Mat_<double>(1, 5) << 0.03468398, 0.44016135, -0.00044522, 0.01792056, -1.15909218); }else{ // ✅ 相机内参矩阵 cameraMatrix: // [[1.05382819e+03 0.00000000e+00 5.72604459e+02] // [0.00000000e+00 1.05425060e+03 4.02170466e+02] // [0.00000000e+00 0.00000000e+00 1.00000000e+00]] // ✅ 畸变系数 distCoeffs: // [[ 0.00272168 -0.04679861 0.00091088 -0.00053716 0.14500516]] this->cameraMatrix = (cv::Mat_<double>(3, 3) << 1.05382819e+03, 0.00000000e+00, 5.72604459e+02, 0.00000000e+00, 1.05425060e+03, 4.02170466e+02, 0.00000000e+00, 0.00000000e+00, 1.00000000e+00); this->distCoeffs = (cv::Mat_<double>(1, 5) << 0.00272168, -0.04679861, 0.00091088, -0.00053716, 0.14500516); } this->squareSize = square_size; // 初始化棋盘格方块尺寸 int major_ver = 0, minor_ver = 0; std::sscanf(CV_VERSION, "%d.%d", &major_ver, &minor_ver); // 使用自适应阈值检测棋盘格角点 // 这里的版本判断是为了兼容OpenCV 4.5+和4.4-的不同函数调用 // 4.5+版本使用findChessboardCornersSB函数，4.4-版本使用findChessboardCorners函数 if (major_ver > 4 || (major_ver == 4 && minor_ver >= 5)){ // OpenCV 4.5+ cvVerPlus45 = true; } } // 棋盘格检测服务回调函数 // 输入：原始图像，输出：位姿向量[x,y,z]（相机坐标系下的棋盘格原点坐标） bool ChessboardDetectionCallback(cv::Mat image, std::vector<float>& vecRt) { if(!ChessboardDetection(image)) { // 执行检测 return false; // 检测失败返回false } // 生成唯一文件名 auto now = std::chrono::system_clock::now(); auto timestamp = std::chrono::duration_cast<std::chrono::milliseconds>( now.time_since_epoch()).count(); std::string filename = "/home/nvidia/YaLongR8/src/bringup/detected_" + std::to_string(timestamp) + ".jpg"; // 保存带角点的图像 if(!cv::imwrite(filename, image)){ std::cerr << "Failed to save image: " << filename << std::endl; } else { std::cout << "Saved image with corners: " << filename << std::endl; } vecRt = PoseCalculation(); // 计算并返回位姿 return true; } // 计算标定板坐标系到相机坐标系的转换矩阵 cv::Point3f transformToCameraFrame(const cv::Point3f& obj_in_board) { cv::Mat R; cv::Rodrigues(rvec, R); // 将旋转向量转为旋转矩阵 cv::Mat pt_board = (cv::Mat_<double>(3,1) << obj_in_board.x, obj_in_board.y, obj_in_board.z); cv::Mat pt_camera = R * pt_board + tvec; return cv::Point3f( static_cast<float>(pt_camera.at<double>(0)), static_cast<float>(pt_camera.at<double>(1)), static_cast<float>(pt_camera.at<double>(2)) ); } private: // 核心检测函数 bool ChessboardDetection(cv::Mat image) { if(image.empty()) { std::cerr << "图像为空！" << std::endl; return false; } const cv::Size boardSize(boardWidth, boardHeight); // 棋盘格内部角点数量（7x7网格） std::vector<cv::Point2f> corners; // 存储检测到的角点坐标 // 转换为灰度图（棋盘格检测需要灰度图像） cv::Mat gray_image; cv::cvtColor(image, gray_image, cv::COLOR_BGR2GRAY); // 使用自适应阈值检测棋盘格角点 bool found = false; if (cvVerPlus45) { ROS_INFO("OpenCV version >= 4.5 detected, using findChessboardCornersSB."); // OpenCV 4.5+ found = cv::findChessboardCornersSB( gray_image, boardSize, corners, cv::CALIB_CB_EXHAUSTIVE | cv::CALIB_CB_ACCURACY ); } else { ROS_INFO("OpenCV version < 4.5 detected, using findChessboardCorners."); // OpenCV 4.4- or earlier fallback // found = cv::findChessboardCorners(gray_image, boardSize, corners); found = cv::findChessboardCorners( gray_image, boardSize, corners, cv::CALIB_CB_ADAPTIVE_THRESH + cv::CALIB_CB_NORMALIZE_IMAGE + cv::CALIB_CB_FAST_CHECK ); } if (!found) { // std::cerr << "未检测到棋盘格！" << std::endl; // cv::imshow("检测结果", image); // 显示原始图像辅助调试 // cv::waitKey(1000); // 显示1秒后自动关闭 // cv::destroyWindow("检测结果"); ROS_ERROR("Chessboard not found!"); return false; } // 亚像素级角点精确化 cv::cornerSubPix( gray_image, corners, cv::Size(11, 11), // 搜索窗口大小 cv::Size(-1, -1), // 死区大小（-1表示无死区） cv::TermCriteria( cv::TermCriteria::EPS + cv::TermCriteria::COUNT, 30, // 最大迭代次数 0.1 // 精度阈值 ) ); // 绘制角点（绿色） cv::drawChessboardCorners(image, boardSize, cv::Mat(corners), found); // 生成三维物体点（假设棋盘格在Z=0平面） std::vector<cv::Point3f> objectPoints; // 棋盘格原点在左上第一个角点处 // for (int i = 0; i < boardSize.height; ++i) { // for (int j = 0; j < boardSize.width; ++j) { // objectPoints.emplace_back( // j * squareSize, // X坐标（沿宽度方向） // i * squareSize, // Y坐标（沿高度方向） // 0 // Z坐标固定为0 // ); // } // } // 棋盘格原点在中心位置 for (int i = 0; i < boardSize.height; ++i) { for (int j = 0; j < boardSize.width; ++j) { objectPoints.emplace_back( (j - (boardSize.width - 1) / 2.0) * squareSize, // X坐标（沿宽度方向） (i - (boardSize.height - 1) / 2.0) * squareSize, // Y坐标（沿高度方向） 0 // Z坐标固定为0 ); } } // 求解PnP问题（透视n点定位） // if (this->rvec.empty() || this->tvec.empty()){ // cv::solvePnP( // objectPoints, // 物体坐标系中的3D点 // corners, // 图像坐标系中的2D点 // cameraMatrix, // 相机内参矩阵 // distCoeffs, // 畸变系数 // rvec, // 输出旋转向量 // tvec, // 输出平移向量 // false, // 使用初始估计（连续帧时提高稳定性） // cv::SOLVEPNP_ITERATIVE // 使用迭代算法 // ); // } else { // cv::solvePnP( // objectPoints, // 物体坐标系中的3D点 // corners, // 图像坐标系中的2D点 // cameraMatrix, // 相机内参矩阵 // distCoeffs, // 畸变系数 // rvec, // 输出旋转向量 // tvec, // 输出平移向量 // true, // 使用初始估计（连续帧时提高稳定性） // cv::SOLVEPNP_ITERATIVE // 使用迭代算法 // ); // } std::vector<int> inliers; cv::solvePnPRansac( objectPoints, corners, cameraMatrix, distCoeffs, rvec, tvec, false, // 不使用初始估计 100, // 最大迭代次数 8.0, // 重投影误差阈值（像素） 0.99, // 置信度 inliers, // 返回的内点索引 cv::SOLVEPNP_ITERATIVE ); return true; } // 位姿计算函数 std::vector<float> PoseCalculation() { // 欧拉角（ZYX） cv::Mat R; cv::Rodrigues(rvec, R); float roll = atan2(R.at<double>(2,1), R.at<double>(2,2)); float pitch = atan2(-R.at<double>(2,0), sqrt(R.at<double>(2,1)R.at<double>(2,1) + R.at<double>(2,2)R.at<double>(2,2))); float yaw = atan2(R.at<double>(1,0), R.at<double>(0,0)); return { static_cast<float>(tvec.at<double>(0)), // X static_cast<float>(tvec.at<double>(1)), // Y static_cast<float>(tvec.at<double>(2)), // Z // static_cast<float>(rvec.at<double>(0)), // Rx // static_cast<float>(rvec.at<double>(1)), // Ry // static_cast<float>(rvec.at<double>(2)) // Rz static_cast<float>(roll), // Roll static_cast<float>(pitch), // Pitch static_cast<float>(yaw) // Yaw }; // // 将旋转向量转换为旋转矩阵 // cv::Mat rotationMatrix; // cv::Rodrigues(rvec, rotationMatrix); // // 计算棋盘格原点在相机坐标系中的坐标 // cv::Mat chessboardOrigin = -rotationMatrix.t() * tvec; // // 转换为float类型向量 // return { // static_cast<float>(chessboardOrigin.at<double>(0)), // X坐标 // static_cast<float>(chessboardOrigin.at<double>(1)), // Y坐标 // static_cast<float>(chessboardOrigin.at<double>(2)) // Z坐标 // }; } };这是啥如何使用

这是一个用于检测棋盘格并计算相机与棋盘格之间位姿关系的C++类。以下是简洁的使用指南： ### 核心功能 1. **棋盘格检测**：自动识别图像中的7×7棋盘格 2. **位姿计算**：计算棋盘格在相机坐标系中的位置和方向 3....

c++使用Eigen3来实现线激光相机y,z值转换为机器人坐标

这需要一个变换矩阵T，它由旋转矩阵R和平移向量t组成。这个变换矩阵描述了如何将相机坐标系下的点转换到机器人基座坐标系。根据引用[2]和[3]，我们可以使用齐次坐标和变换矩阵进行转换。设相机坐标系下的点为$P_c...

#include <ros/ros.h> #include <image_transport/image_transport.h> #include <cv_bridge/cv_bridge.h> #include <sensor_msgs/image_encodings.h> #include <opencv2/imgproc/imgproc.hpp> #include <opencv2/highgui/highgui.hpp> #include <stdio.h> #include <stdlib.h> #include "opencv2/opencv.hpp" #include <moveit/move_group_interface/move_group_interface.h> #include <tf/transform_listener.h> #include <moveit/planning_scene_interface/planning_scene_interface.h> #include <moveit/robot_trajectory/robot_trajectory.h> #include "myself_pkg/uart.h" #include <sys/stat.h> #include <cmath> #include <xarm_driver.h> #include <thread> // 引入线程库 #include <atomic> static const std::string OPENCV_WINDOW = "Image window"; #define M_PI 3.14159265358979323846 tf::Vector3 obj_camera_frame1, obj_robot_frame; int flag_start=1; ros::Subscriber rgb_sub; // 相机内参 double fx = 953.4568; // x轴方向的焦距 double fy = 949.837; // y轴方向的焦距 double cx = 658.66659; // x轴方向的光学中心 double cy = 366.82704; // y轴方向的光学中心 // 物体高度 double objectHeight = 0.34432666; // 假设物体的高度为1米 double k1 = 0.133507; // double k2 =-0.213178; // double p1 = 0.006242; // double p2 = 0.005494; // /* double fx = 1084.54479; // x轴方向的焦距 double fy = 950.11576; // y轴方向的焦距 double cx = 642.85519; // x轴方向的光学中心 double cy = 354.52482; // y轴方向的光学中心 // 物体高度 double objectHeight = 0.34432666; // 假设物体的高度为1米 double k1 = 0.141430; // double k2 =-0.384089; // double p1 = 0.003167; // double p2 = 0.002440; // */ cv::Mat K = (cv::Mat_<double>(3, 3) << fx, 0, cx, 0, fy, cy, 0, 0, 1); // 构造畸变参数向量 cv::Mat distCoeffs = (cv::Mat_<double>(5, 1) << k1, k2, p1, p2, 0); // 待校正的像素坐标 cv::Point2d pixelPoint(640, 360); // 假设像素坐标为(320, 240) cv::Mat cameraPointMat = (cv::Mat_<double>(3, 1) << 0, 0,0); /** // 定义一个结构体表示四元数，用于三维空间中的旋转表示 struct Quaternion { double w; // 四元数的实部 double x; // 四元数的虚部 x double y; // 四元数的虚部 y double z; // 四元数的虚部 z }; Quaternion eulerToQuaternion(double roll, double pitch, double yaw) { // 计算半角 double cy = cos(yaw * 0.5); double sy = sin(yaw * 0.5); double cp = cos(pitch * 0.5); double sp = sin(pitch * 0.5); double cr = cos(roll * 0.5); double sr = sin(roll * 0.5); Quaternion q; // 根据欧拉角到四元数的转换公式计算四元数的实部 q.w = cr * cp * cy + sr * sp * sy; // 根据欧拉角到四元数的转换公式计算四元数的虚部 x q.x = sr * cp * cy - cr * sp * sy; // 根据欧拉角到四元数的转换公式计算四元数的虚部 y q.y = cr * sp * cy + sr * cp * sy; // 根据欧拉角到四元数的转换公式计算四元数的虚部 z q.z = cr * cp * sy - sr * sp * cy; return q; } */ int move_lineb_test(xarm_msgs::Move srv, ros::ServiceClient client, float x_mm0, float y_mm0, float z_mm0, double roll0, double pitch0, double yaw0, float x_mm1, float y_mm1, float z_mm1, double roll1, double pitch1, double yaw1, float x_mm2, float y_mm2, float z_mm2, double roll2, double pitch2, double yaw2, float x_mm3, float y_mm3, float z_mm3, double roll3, double pitch3, double yaw3, float x_mm4, float y_mm4, float z_mm4, double roll4, double pitch4, double yaw4); void control_suction_during_move(float x_mm4, float y_mm4, float z_mm4); // 异步吸盘控制 auto suction_control = [](int speed){ std::thread([speed](){ for(int i=0; i<1; i++){ writeSpeed(speed); std::this_thread::sleep_for(std::chrono::milliseconds(800)); } }).detach(); // 分离线程 }; class XArmAPItest { ros::NodeHandle nh_; image_transport::ImageTransport it_; image_transport::Subscriber image_sub_; tf::TransformListener listener_; tf::StampedTransform camera_to_robot_; public: XArmAPItest() : it_(nh_) { moveit::planning_interface::MoveGroupInterface arm("xarm7"); sleep(0.5); moveit::planning_interface::PlanningSceneInterface planning_scene_interface; sleep(0.5); //异步任务处理器，防阻塞 ros::AsyncSpinner spinner(1); spinner.start(); // 创建一个新的障碍物消息 moveit_msgs::CollisionObject collision_object; collision_object.header.frame_id = "world"; // 设置障碍物的参考坐标系，通常为世界坐标系 // 设置障碍物的 ID collision_object.id = "table"; // 定义障碍物的形状和尺寸 shape_msgs::SolidPrimitive primitive; primitive.type = primitive.BOX; primitive.dimensions.resize(3); primitive.dimensions[0] = 2.0; // 长 primitive.dimensions[1] = 2.0; // 宽 primitive.dimensions[2] = 0.1; // 高 // 定义障碍物的姿态 geometry_msgs::Pose obstacle_pose; obstacle_pose.orientation.w = 1.0; // 默认姿态为单位四元数 obstacle_pose.position.x = 0.0; // x 位置 obstacle_pose.position.y = 0.0; // y 位置 obstacle_pose.position.z = -0.06; // z 位置 // 将障碍物的形状和姿态添加到障碍物消息中 collision_object.primitives.push_back(primitive); collision_object.primitive_poses.push_back(obstacle_pose); // 设置操作类型为添加障碍物 collision_object.operation = collision_object.ADD; // 发送障碍物消息到规划场景 moveit_msgs::PlanningScene planning_scene; planning_scene.world.collision_objects.push_back(collision_object); planning_scene.is_diff = true; planning_scene_interface.applyPlanningScene(planning_scene); // 应用障碍物到规划场景 ROS_INFO("Obstacle added"); /** //回初位置 arm.setNamedTarget("home");//设置目标 arm.move();//执行 sleep(0.5); double targetPose[7] = {-0.166690,0.00000, -0.076904,1.173601, 0.015010,1.21220,-0.260379}; std::vector<double> joint_group_positions(7); joint_group_positions[0] = targetPose[0]; joint_group_positions[1] = targetPose[1]; joint_group_positions[2] = targetPose[2]; joint_group_positions[3] = targetPose[3]; joint_group_positions[4] = targetPose[4]; joint_group_positions[5] = targetPose[5]; joint_group_positions[6] = targetPose[6]; arm.setJointValueTarget(joint_group_positions); arm.move(); sleep(0.5); */ try { listener_.waitForTransform("link_base", "camera_link", ros::Time(0), ros::Duration(50.0)); } catch (tf::TransformException &ex) { ROS_ERROR("[adventure_tf]: (wait) %s", ex.what()); ros::Duration(1.0).sleep(); } try { listener_.lookupTransform("link_base", "camera_link", ros::Time(0), camera_to_robot_); tf::Vector3 translation = camera_to_robot_.getOrigin(); objectHeight=translation.getZ(); std::cout << objectHeight << std::endl; } catch (tf::TransformException &ex) { ROS_ERROR("[adventure_tf]: (lookup) %s", ex.what()); } // 订阅相机图像 image_sub_ = it_.subscribe("/camera/color/image_raw", 1, &XArmAPItest::Cam_RGB_Callback, this); sleep(1); } /** // 封装四元数转欧拉角的函数 std::tuple<double, double, double> quaternionToEuler(const geometry_msgs::Quaternion& q) { double x = q.x; double y = q.y; double z = q.z; double w = q.w; // 计算绕 x 轴旋转的弧度（roll） double sinr_cosp = 2 * (w * x + y * z); double cosr_cosp = 1 - 2 * (x * x + y * y); double roll = std::atan2(sinr_cosp, cosr_cosp); // 计算绕 y 轴旋转的弧度（pitch） double sinp = 2 * (w * y - z * x); double pitch; if (std::abs(sinp) >= 1) pitch = std::copysign(M_PI / 2, sinp); // 使用 90 度避免数值问题 else pitch = std::asin(sinp); // 计算绕 z 轴旋转的弧度（yaw） double siny_cosp = 2 * (w * z + x * y); double cosy_cosp = 1 - 2 * (y * y + z * z); double yaw = std::atan2(siny_cosp, cosy_cosp); return std::make_tuple(roll, pitch, yaw); } / void Grasping(double a, double b, double z, double Angle) { std::cout << "Grasping" << std::endl; // 输出提示信息 moveit::planning_interface::MoveGroupInterface arm("xarm7"); ros::AsyncSpinner spinner(1); spinner.start(); std::string end_effector_link = arm.getEndEffectorLink(); std::string reference_frame = "link_base"; arm.setPoseReferenceFrame(reference_frame); // 声明一个变量用于存储机械臂当前的位姿信息 geometry_msgs::PoseStamped homePose; // 获取机械臂末端执行器当前的位姿并赋值给 homePose 变量 homePose = arm.getCurrentPose(); sleep(1); writeSpeed(1); writeSpeed(1); writeSpeed(1); sleep(0.8); /第1次抓取/ / if(flag_start==1) { //移动到抓取方块上方 flag_start=2; //将初始位姿加入路点列表 //waypoints.push_back(target_pose); geometry_msgs::Pose target_pose; target_pose.position.x = x; target_pose.position.y = y; target_pose.position.z = 0; double roll = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch = 0; // 绕 y 轴旋转的弧度 double yaw = 0; // 绕 z 轴旋转的弧度 // 将目标姿态的位置坐标从米转换为毫米 float x_mm = static_cast<float>(target_pose.position.x 1000); float y_mm = static_cast<float>(target_pose.position.y * 1000); float z_mm = static_cast<float>(target_pose.position.z * 1000); } / double x = a + 0.0001-0.0041; double y = b + 0.0009+0.0082+0.001; /抓取***/ if(flag_start>=1&&flag_start<35) { //抓取物块 geometry_msgs::Pose target_pose0; target_pose0 = arm.getCurrentPose(end_effector_link).pose; target_pose0.position.z = 0+0.14+0.013; target_pose0.position.x = x; target_pose0.position.y = y; double roll0 = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch0 =0M_PI/180; // 绕 y 轴旋转的弧度 double yaw0 = 0; // 绕 z 轴旋转的弧度 // 将目标姿态的位置坐标从米转换为毫米 float x_mm0 = static_cast<float>(target_pose0.position.x * 1000); float y_mm0 = static_cast<float>(target_pose0.position.y * 1000); float z_mm0 = static_cast<float>(target_pose0.position.z * 1000); ROS_INFO("zuobiao0: %f, %f, %f", x_mm0, y_mm0, z_mm0); // 定位抓取物块 geometry_msgs::Pose target_pose1; target_pose1 = arm.getCurrentPose(end_effector_link).pose; target_pose1.position.x = x; target_pose1.position.y = y; target_pose1.position.z = 0.09+0.0003+0.01246-0.01; double roll1 = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch1 =0M_PI/180; // 绕 y 轴旋转的弧度 double yaw1 = 0; // 绕 z 轴旋转的弧度 // 将目标姿态的位置坐标从米转换为毫米 float x_mm1 = static_cast<float>(target_pose1.position.x * 1000); float y_mm1 = static_cast<float>(target_pose1.position.y * 1000); float z_mm1 = static_cast<float>(target_pose1.position.z * 1000); ROS_INFO("zuobiao1: %f, %f, %f", x_mm1, y_mm1, z_mm1); //抓取物块 geometry_msgs::Pose target_pose2; target_pose2 = arm.getCurrentPose(end_effector_link).pose; target_pose2.position.z = 0+0.14+0.013; target_pose2.position.x = x; target_pose2.position.y = y; double roll2 = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch2 =0M_PI/180; // 绕 y 轴旋转的弧度 double yaw2 = 0; // 绕 z 轴旋转的弧度 // 将目标姿态的位置坐标从米转换为毫米 float x_mm2 = static_cast<float>(target_pose2.position.x * 1000); float y_mm2 = static_cast<float>(target_pose2.position.y * 1000); float z_mm2 = static_cast<float>(target_pose2.position.z * 1000); ROS_INFO("zuobiao2: %f, %f, %f", x_mm2, y_mm2, z_mm2); geometry_msgs::Pose target_pose3; target_pose3 = arm.getCurrentPose(end_effector_link).pose; //放置物块 target_pose3.position.z= 0.14+0.013; double roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 double yaw3 = Angle*M_PI/180; // 绕 z 轴旋转的弧度 /****************************zi色*/ if(flag_start==1){ target_pose3.position.x = 0.38578-0.023+0.002; target_pose3.position.y = 0.217467-0.052;} if(flag_start==2){ target_pose3.position.x = 0.38578+0.06-0.023; target_pose3.position.y = 0.217467+0.02-0.052;} if(flag_start==3){ target_pose3.position.x = 0.38578+0.08-0.023; target_pose3.position.y = 0.217467-0.052;} /**********橙色*/ if(flag_start==4){ target_pose3.position.x = 0.38578+0.01-0.023; target_pose3.position.y = 0.217467+0.052-0.055;} if(flag_start==5){ target_pose3.position.x = 0.38578+0.05-0.023; target_pose3.position.y = 0.217467+0.052-0.055;} if(flag_start==6){ target_pose3.position.x = 0.38578+0.09-0.023; target_pose3.position.y = 0.217467+0.052-0.055;} /*******hong色 */ if(flag_start==7){ target_pose3.position.x = 0.38578-0.023; target_pose3.position.y = 0.217467+0.11-0.055+0.002;} if(flag_start==8){ target_pose3.position.x = 0.38578+0.02-0.023-0.002; target_pose3.position.y = 0.217467+0.11-0.055+0.002;} if(flag_start==9){ target_pose3.position.x = 0.38578+0.04-0.023-0.002; target_pose3.position.y = 0.217467+0.11-0.055;} /**********************huang色*/ if(flag_start==10){ target_pose3.position.x = 0.38578-0.023+0.002; target_pose3.position.y = 0.217467+0.18+-0.05-0.001;} if(flag_start==11){ target_pose3.position.x = 0.38578+0.06-0.023-0.002; target_pose3.position.y = 0.217467+0.16-0.05;} if(flag_start==12){ target_pose3.position.x = 0.38578+0.08-0.023; target_pose3.position.y = 0.217467+0.18-0.05;} if(flag_start==13){ target_pose3.position.x = 0.38578+0.14-0.023-0.001; target_pose3.position.y = 0.217467+0.16-0.05;} /*****hei色* */ if(flag_start==14){ target_pose3.position.x = 0.38578+0.06-0.023; target_pose3.position.y = 0.217467+0.10-0.05;} if(flag_start==15){ target_pose3.position.x = 0.38578+0.08-0.023; target_pose3.position.y = 0.217467+0.14-0.05-0.001;} /***************************zi色* */ if(flag_start==16){ target_pose3.position.x = 0.38578+0.12-0.023-0.001; target_pose3.position.y = 0.217467+0.08-0.0575;} /*********************lan色* / if(flag_start==17){ target_pose3.position.x = 0.38578+0.11-0.023+0.001; target_pose3.position.y = 0.217467+0.12-0.05-0.001; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = AngleM_PI/180 + M_PI/2; // 绕 z 轴旋转的弧度 } if(flag_start==18){ target_pose3.position.x = 0.38578+0.11+0.03-0.02; target_pose3.position.y = 0.217467+0.09-0.05;} if(flag_start==19){ target_pose3.position.x = 0.38578+0.11+0.03+0.04-0.02-0.002; target_pose3.position.y = 0.217467+0.09-0.05+0.001;} if(flag_start==20){ target_pose3.position.x = 0.38578+0.11+0.03+0.04+0.02-0.02-0.02-0.07; target_pose3.position.y = 0.217467+0.09+0.02-0.05+0.002;} /*************************ceng色* */ if(flag_start==21){ target_pose3.position.x = 0.38578+0.15-0.02-0.002; target_pose3.position.y = 0.217467+0.13-0.045-0.002;} /*************lv色 / if(flag_start==22){ target_pose3.position.x = 0.38578+0.12-0.02-0.004; target_pose3.position.y = 0.217467+0.03-0.05;} if(flag_start==23){ target_pose3.position.x = 0.38578+0.16-0.02-0.004-0.0005; target_pose3.position.y = 0.217467+0.03-0.05;} if(flag_start==24){ target_pose3.position.x = 0.38578+0.20-0.02-0.004-0.0005; target_pose3.position.y = 0.217467+0.03-0.05;} if(flag_start==25){ target_pose3.position.x = 0.38578+0.17-0.02-0.004-0.0001; target_pose3.position.y = 0.217467+0.11+0.06-0.05-0.005; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = AngleM_PI/180 - M_PI/2; // 绕 z 轴旋转的弧度 } /*********************huang色 */ if(flag_start==26){ target_pose3.position.x = 0.38578+0.20-0.022; target_pose3.position.y = 0.217467+0.11+0.08-0.05-0.004-0.004;} /*********************hong色 / if(flag_start==27){ target_pose3.position.x = 0.38578+0.17-0.021; target_pose3.position.y = 0.217467-0.05; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = AngleM_PI/180 - M_PI/2; // 绕 z 轴旋转的弧度 } if(flag_start==28){ target_pose3.position.x = 0.38578+0.17-0.021; target_pose3.position.y = 0.217467+0.06-0.05; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = Angle*M_PI/180 - M_PI/2; // 绕 z 轴旋转的弧度 } /***************************hei色* */ if(flag_start==29){ target_pose3.position.x = 0.38578+0.22-0.025+0.002; target_pose3.position.y = 0.217467+0.10+0.04-0.05;} if(flag_start==30){ target_pose3.position.x = 0.38578+0.24-0.025; target_pose3.position.y = 0.217467+0.18-0.05;} /***********************ceng色* */ if(flag_start==31){ target_pose3.position.x = 0.38578+0.23-0.02-0.002-0.0002; target_pose3.position.y = 0.217467+0.07-0.045-0.002-0.002;} /***********************lan色* / if(flag_start==32){ target_pose3.position.x = 0.38578+0.23-0.024; target_pose3.position.y = 0.217467+0.02-0.051-0.0001; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = AngleM_PI/180 + M_PI/2; // 绕 z 轴旋转的弧度 } /***********************zi色* */ if(flag_start==33){ target_pose3.position.x = 0.38578+0.26-0.024; target_pose3.position.y = 0.217467-0.051+0.001;} /***********************lv色* / if(flag_start==34){ target_pose3.position.x = 0.38578+0.25-0.024; target_pose3.position.y = 0.217467+0.10-0.051+0.001; target_pose3.position.z= 0.14+0.013+0.002; roll3 = 180M_PI/180; // 绕 x 轴旋转的弧度 pitch3 =0M_PI/180; // 绕 y 轴旋转的弧度 yaw3 = AngleM_PI/180 - M_PI/2; // 绕 z 轴旋转的弧度 } // 将目标姿态的位置坐标从米转换为毫米 float x_mm3 = static_cast<float>(target_pose3.position.x * 1000); float y_mm3 = static_cast<float>(target_pose3.position.y * 1000); float z_mm3 = static_cast<float>(target_pose3.position.z * 1000); ROS_INFO("zuobiao3: %f, %f, %f", x_mm3, y_mm3, z_mm3); geometry_msgs::Pose target_pose4; target_pose4.position.z= 0.109+0.01236-0.0078+0.005; double roll4 = 180M_PI/180; // 绕 x 轴旋转的弧度 double pitch4 =0M_PI/180; // 绕 y 轴旋转的弧度 double yaw4 = yaw3; // 绕 z 轴旋转的弧度 float x_mm4 = static_cast<float>(target_pose3.position.x * 1000); float y_mm4 = static_cast<float>(target_pose3.position.y * 1000); float z_mm4 = static_cast<float>(target_pose4.position.z * 1000); ROS_INFO("zuobiao4: %f, %f, %f", x_mm4, y_mm4, z_mm4); // 调用 move_lineb_test 函数并传递坐标 ros::NodeHandle nh; ros::ServiceClient move_lineb_client_ = nh.serviceClient<xarm_msgs::Move>("/xarm/move_lineb"); xarm_msgs::Move move_srv_; if(move_lineb_test(move_srv_, move_lineb_client_, x_mm0, y_mm0, z_mm0, roll0, pitch0, yaw0, x_mm1, y_mm1, z_mm1, roll1, pitch1, yaw1, x_mm2, y_mm2, z_mm2, roll2, pitch2, yaw2, x_mm3, y_mm3, z_mm3, roll3, pitch3, yaw3, x_mm4, y_mm4, z_mm4, roll4, pitch4, yaw4) == 1) return; control_suction_during_move(x_mm4, y_mm4, z_mm4); flag_start++; } else {ros::shutdown(); } } void Cam_RGB_Callback(const sensor_msgs::ImageConstPtr &msg)// 摄像头回调函数 { using namespace cv; image_sub_.shutdown(); // 定义一个cv_bridge指针 cv_bridge::CvImagePtr cv_ptr; try { // 将ROS图像转换为OpenCV图像 cv_ptr = cv_bridge::toCvCopy(msg, sensor_msgs::image_encodings::BGR8); } catch (cv_bridge::Exception &e) { ROS_ERROR("cv_bridge exception:%s", e.what()); } // 获取原始图像 Mat imgOriginal = cv_ptr->image; // 定义亮度增强因子 double brightness_scale = 1.8; // 应用亮度增强 Mat brightened; imgOriginal.convertTo(brightened, -1, brightness_scale); // 图像预处理：高斯模糊 Mat blurred; GaussianBlur(brightened, blurred, Size(5, 5), 0); Mat hsv; cvtColor(blurred, hsv, cv::COLOR_BGR2HSV); // 将原始图像转换为HSV图像 // 分离HSV通道 std::vector<Mat> hsv_channels; split(hsv, hsv_channels); // 增强饱和度（S 通道） double saturation_scale = 1.5; // 饱和度增强因子，可以根据实际情况调整 hsv_channels[1].convertTo(hsv_channels[1], -1, saturation_scale); // 合并通道 merge(hsv_channels, hsv); Mat mask_red, mask_green,mask_blue,mask_orange,mask_brown,mask_yellow,mask_purple; // inRange(blurred, cv::Scalar(0, 0, 130), cv::Scalar(255, 108, 226), mask_red); inRange(blurred, cv::Scalar(0, 0, 144), cv::Scalar(252, 111, 203), mask_red); inRange(blurred, cv::Scalar(0, 112, 174), cv::Scalar(159, 139, 237), mask_orange);//BGR inRange(blurred, cv::Scalar(0, 0, 0), cv::Scalar(123, 112, 108), mask_brown); inRange(blurred, cv::Scalar(0, 127, 97), cv::Scalar(157, 255, 136), mask_green); inRange(blurred, cv::Scalar(156, 119, 0), cv::Scalar(218, 146, 102), mask_blue); inRange(blurred, cv::Scalar(0, 142, 152), cv::Scalar(150, 255, 255), mask_yellow); inRange(hsv, cv::Scalar(99, 33, 120), cv::Scalar(134, 146, 181), mask_purple); Mat kernel = getStructuringElement(MORPH_ELLIPSE, Size(5, 5));// 形态学操作的内核大小 //dilate(mask_red, mask_red, kernel); //dilate(mask_green, mask_green, kernel); erode(mask_red, mask_red, kernel);//腐蚀 erode(mask_green, mask_green, kernel); erode(mask_blue, mask_blue, kernel); erode(mask_yellow, mask_yellow, kernel); erode(mask_orange, mask_orange, kernel); erode(mask_purple, mask_purple, kernel); erode(mask_brown, mask_brown, kernel); //erode(mask_purple, mask_purple, kernel); dilate(mask_orange, mask_orange, kernel); dilate(mask_brown, mask_brown, kernel); //imshow("green", mask_green);//显示原始图像 //获取储存不同颜色的灰度图 std::vector<std::vector<cv::Point>> contours_red; std::vector<std::vector<cv::Point>> contours_red_output; cv::findContours(mask_red,contours_red, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_orange; std::vector<std::vector<cv::Point>> contours_orange_output; cv::findContours(mask_orange,contours_orange, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_brown; std::vector<std::vector<cv::Point>> contours_brown_output; cv::findContours(mask_brown,contours_brown, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_green; std::vector<std::vector<cv::Point>> contours_green_output; cv::findContours(mask_green,contours_green, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_blue; std::vector<std::vector<cv::Point>> contours_blue_output; cv::findContours(mask_blue,contours_blue, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_yellow; std::vector<std::vector<cv::Point>> contours_yellow_output; cv::findContours(mask_yellow,contours_yellow, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); std::vector<std::vector<cv::Point>> contours_purple; std::vector<std::vector<cv::Point>> contours_purple_output; cv::findContours(mask_purple,contours_purple, cv::RETR_EXTERNAL, cv::CHAIN_APPROX_NONE); //drawContours(blurred, contours, -1, Scalar(0, 255, 0), 3);//轮廓 /************红色/ for (size_t i = 0; i < contours_red.size(); i++) { std::vector<cv::Point>& contour_red = contours_red[i]; if (contour_red.size()>200) { // 检查轮廓是否为空 contours_red_output.push_back(contours_red[i]); } else { ROS_INFO("No red contours found."); } } /橙色/ for (size_t i = 0; i < contours_orange.size(); i++) { std::vector<cv::Point>& contour_orange = contours_orange[i]; if (contour_orange.size()>200) { // 检查轮廓是否为空 contours_orange_output.push_back(contours_orange[i]); } else { ROS_INFO("No orange contours found."); } } /棕色/ for (size_t i = 0; i < contours_brown.size(); i++) { std::vector<cv::Point>& contour_brown = contours_brown[i]; if (contour_brown.size()>200) { // 检查轮廓是否为空 contours_brown_output.push_back(contours_brown[i]); } else { ROS_INFO("No brown contours found."); } } /绿色/ for (size_t i = 0; i < contours_green.size(); i++) { std::vector<cv::Point>& contour_green = contours_green[i]; if (contour_green.size()>200) { // 检查轮廓是否为空 contours_green_output.push_back(contours_green[i]); } else { ROS_INFO("No green contours found."); } } /蓝色/ for (size_t i = 0; i < contours_blue.size(); i++) { std::vector<cv::Point>& contour_blue = contours_blue[i]; if (contour_blue.size()>200) { // 检查轮廓是否为空 contours_blue_output.push_back(contours_blue[i]); } else { ROS_INFO("No blue contours found."); } } /黄色/ for (size_t i = 0; i < contours_yellow.size(); i++) { std::vector<cv::Point>& contour_yellow = contours_yellow[i]; if (contour_yellow.size()>200) { // 检查轮廓是否为空 contours_yellow_output.push_back(contours_yellow[i]); } else { ROS_INFO("No yellow contours found."); } } /紫色/ // 存储每个轮廓的面积及其索引 std::vector<std::pair<double, size_t>> area_index_pairs; for (size_t i = 0; i < contours_purple.size(); ++i) { double area = cv::contourArea(contours_purple[i]); area_index_pairs.emplace_back(area, i); } // 按面积从大到小排序 std::sort(area_index_pairs.begin(), area_index_pairs.end(), [](const std::pair<double, size_t>& a, const std::pair<double, size_t>& b) { return a.first > b.first; }); // 处理面积最大的5个轮廓 size_t count = std::min<size_t>(5, area_index_pairs.size()); if (count == 0) { ROS_INFO("No contours found."); } for (size_t i = 0; i < count; ++i) { size_t index = area_index_pairs[i].second; contours_purple_output.push_back(contours_purple[index]); } //传递颜色灰度图像 Camera_TO_Robot_Process_YP(contours_purple_output,0,3); Camera_TO_Robot_Process_RO(contours_orange_output,0,3); Camera_TO_Robot_Process_RO(contours_red_output,0,3); Camera_TO_Robot_Process_YP(contours_yellow_output,0,4); Camera_TO_Robot_Process_brown(contours_brown_output,0,2); Camera_TO_Robot_Process_YP(contours_purple_output,3,4); Camera_TO_Robot_Process_GB(contours_blue_output,0,4); Camera_TO_Robot_Process_RO(contours_orange_output,3,4); Camera_TO_Robot_Process_GB(contours_green_output,0,4); Camera_TO_Robot_Process_YP(contours_yellow_output,4,5); Camera_TO_Robot_Process_RO(contours_red_output,3,5); Camera_TO_Robot_Process_brown(contours_brown_output,2,4); Camera_TO_Robot_Process_RO(contours_orange_output,4,5); Camera_TO_Robot_Process_GB(contours_blue_output,4,5); Camera_TO_Robot_Process_YP(contours_purple_output,4,5); Camera_TO_Robot_Process_GB(contours_green_output,4,5); // imshow("imgOriginal", imgOriginal); std::cout << "success!" << std::endl; ros::shutdown(); } //红色和橙色 void Camera_TO_Robot_Process_RO(const std::vector<std::vector<cv::Point>>& contours, int start_number, int end_number) { std::cout << "Red or Orange" << std::endl; cv::Point2f center; for (start_number; start_number < end_number; start_number++) { const std::vector<cv::Point>& contour = contours[start_number]; if (contour.size()>300) // 检查轮廓是否为空 { // 获取最小外接圆 // float radius; // cv::minEnclosingCircle(contours[i], center, radius); // 获取最小外接矩形 cv::RotatedRect minRect = cv::minAreaRect(contours[start_number]); center = minRect.center; if(center.y<60) { center.y=60; } if(center.x<82) { center.x=82; } if(center.y>=60&¢er.x>=82) { cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x, center.y,1); //cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x1.2-91.0, center.y1.275-74.5,1); std::cout << pixelPointMat << std::endl; // pixelPointMat = (cv::Mat_<double>(3, 1) << 640, 360, 1); cameraPointMat = objectHeight * K.inv() * pixelPointMat; // std::cout << "相机坐标系下的三维坐标：" << std::endl; // std::cout << cameraPointMat << std::endl; obj_camera_frame1.setX(cameraPointMat.at<double>(0,0)); obj_camera_frame1.setY(cameraPointMat.at<double>(1,0)); obj_camera_frame1.setZ(cameraPointMat.at<double>(2,0)); obj_robot_frame = camera_to_robot_ * obj_camera_frame1; // ros::shutdown(); //std::cout << "11111" << std::endl; /****** 获取矩形的角度/ double angle = minRect.angle; cv::Size2f size = minRect.size; double width = size.width; double height = size.height; //放平角度 if (width < height) { angle += 90; } // 输出角度 // std::cout << "Contour #" << start_number << " angle: " << angle << std::endl; / 获取矩形的角度/ Grasping(obj_robot_frame.getX(),obj_robot_frame.getY(),obj_robot_frame.getZ(),angle); } } else if(contour.size()>0&&contour.size()<100) { std::cout << "没有红橙像素坐标：" << std::endl; } } } //棕色 void Camera_TO_Robot_Process_brown(const std::vector<std::vector<cv::Point>>& contours, int start_number, int end_number) { std::cout << "Brown start" << std::endl; cv::Point2f center; for (start_number; start_number < end_number; start_number++) { const std::vector<cv::Point>& contour = contours[start_number]; if (contour.size()>300) // 检查轮廓是否为空 { // 获取最小外接圆 float radius; cv::minEnclosingCircle(contours[start_number], center, radius); / 获取矩形的角度*/ // 获取最小外接矩形 cv::RotatedRect minRect = cv::minAreaRect(contours[start_number]); double angle2 = minRect.angle; cv::Size2f size = minRect.size; double width = size.width; double height = size.height; / 获取矩形的角度***********/ double epsilon = 0.1 * cv::arcLength(contours[start_number], true); std::cout << "1111" << std::endl; std::vector<int> lenth(8); std::vector<cv::Point2f> approx; cv::Point2f pt; cv::Point2f pt1; cv::Point2f pt0; cv::approxPolyDP(contours[start_number], approx, epsilon, true); pt0 = approx[0]; if (approx.size() == 3) { for (size_t j = 0; j < approx.size(); j++) { pt= approx[j]; pt1= approx[j+1]; lenth[j]=(pt.x-pt1.x)(pt.x-pt1.x)+(pt.y-pt1.y)(pt.y-pt1.y); if(j==2) { lenth[j]=(pt.x-pt0.x)(pt.x-pt0.x)+(pt.y-pt0.y)(pt.y-pt0.y); } } if ( lenth[0]>lenth[1]&&lenth[0]>lenth[2]) { if(width > height){ switch (start_number) { case 0: angle2 = angle2; break; case 1: angle2 -= 90; break; case 2: angle2 = angle2; break; case 3: angle2 -= 90; break; } } else if(width < height){ switch (start_number) { case 0: angle2 -= 90; break; case 1: angle2 += 180; break; case 2: angle2 -= 90; break; case 3: angle2 += 180; break; } } } else if ( lenth[2]>lenth[1]&&lenth[2]>lenth[0]) { if(width > height){ switch (start_number) { case 0: angle2 += 180; break; case 1: angle2 += 90; break; case 2: angle2 += 180; break; case 3: angle2 += 90; break; } } else if(width < height){ switch (start_number) { case 0: angle2 += 90; break; case 1: angle2 = angle2; break; case 2: angle2 += 90; break; case 3: angle2 = angle2; break; } } } } // center = minRect.center; if(center.y<60) { center.y=60; } if(center.x<82) { center.x=82; } if(center.y>=60&¢er.x>=82) { //cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x1.2-91.5, center.y1.275-74.5,1); //cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x1.2-92, center.y1.285-76.5,1); cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x, center.y,1); std::cout << pixelPointMat << std::endl; // pixelPointMat = (cv::Mat_<double>(3, 1) << 640, 360, 1); cameraPointMat = objectHeight * K.inv() * pixelPointMat; // std::cout << "相机坐标系下的三维坐标：" << std::endl; // std::cout << cameraPointMat << std::endl; obj_camera_frame1.setX(cameraPointMat.at<double>(0,0)); obj_camera_frame1.setY(cameraPointMat.at<double>(1,0)); obj_camera_frame1.setZ(cameraPointMat.at<double>(2,0)); obj_robot_frame = camera_to_robot_ * obj_camera_frame1; if(angle2>176) angle2=176; if(angle2<-176) angle2=-176; std::cout << " angle2 :" << angle2 << std::endl; Grasping(obj_robot_frame.getX(),obj_robot_frame.getY(),obj_robot_frame.getZ(),angle2); } } else if(contour.size()>0&&contour.size()<100) { std::cout << "没有棕色像素坐标：" << std::endl; }} } //紫色和黄色 void Camera_TO_Robot_Process_YP(const std::vector<std::vector<cv::Point>>& contours, int start_number, int end_number) { std::cout << "Yellow or Purple start" << std::endl; double angle2=0; for (start_number; start_number < end_number; start_number++) { const std::vector<cv::Point>& contour = contours[start_number]; if (contour.size()>300) // 检查轮廓是否为空 { /****** 获取矩形的角度*/ cv::RotatedRect minRect = cv::minAreaRect(contours[start_number]); angle2 = minRect.angle; cv::Size2f size = minRect.size; double width = size.width; double height = size.height; / 获取矩形的角度***********/ cv::Point2f center; cv::Point2f center2; double epsilon = 0.04 * cv::arcLength(contours[start_number], true); // 多边形逼近 std::vector<int> lenth(6); std::vector<cv::Point2f> approx; cv::Point2f pt; cv::Point2f pt1; cv::Point2f pt0; cv::approxPolyDP(contours[start_number], approx, epsilon, true); // std::cout <<" lenth[i] "<< approx.size() << std::endl; pt0 = approx[0]; if (approx.size() == 6) { for (size_t i = 0; i < approx.size(); i++) { pt= approx[i]; pt1= approx[i+1]; lenth[i]=(pt.x-pt1.x)(pt.x-pt1.x)+(pt.y-pt1.y)(pt.y-pt1.y); if(i==5) lenth[i]=(pt.x-pt0.x)(pt.x-pt0.x)+(pt.y-pt0.y)(pt.y-pt0.y); } /************purple/ if ( lenth[5]>lenth[1]&&lenth[5]>lenth[0]&&lenth[5]>lenth[2]&&lenth[5]>lenth[3]&&lenth[5]>lenth[4]) { // center2 = (approx[1]+approx[2])/2; // center = (approx[5]+center2)/2; center = (approx[5]+approx[2])/2; if(width > height){ switch (start_number) { case 0: angle2 -= 90; break; case 1: angle2 += 90; break; case 2: angle2 -= 90; break; case 3: angle2 += 90; break; case 4: angle2 = angle2; break; }} else if(width < height){ switch (start_number) { case 0: angle2 += 180; break; case 1: angle2 = angle2; break; case 2: angle2 += 180; break; case 3: angle2 = angle2; break; case 4: angle2 -= 90; break; }} std::cout << "555555" << std::endl; } /purple/ else if ( lenth[1]>lenth[2]&&lenth[1]>lenth[0]&&lenth[1]>lenth[5]&&lenth[1]>lenth[3]&&lenth[1]>lenth[4]) { //center2 = (approx[3]+approx[4])/2; //center = (approx[1]+center2)/2; center = (approx[1]+approx[4])/2; if(width > height){ if(approx[1].y<approx[0].y&&approx[1].y<approx[2].y&&approx[1].y<approx[3].y&&approx[1].y<approx[4].y&&approx[1].y<approx[5].y){ switch (start_number) { case 0: angle2 += 90; break; case 1: angle2 -= 90; break; case 2: angle2 += 90; break; case 3: angle2 -= 90; break; case 4: angle2 += 180; break; }} else{ switch (start_number) { case 0: angle2 -= 90; break; case 1: angle2 += 90; break; case 2: angle2 -= 90; break; case 3: angle2 += 90; break; case 4: angle2 = angle2; break; } } } else if(width < height) { switch (start_number) { case 0: angle2 = angle2; break; case 1: angle2 += 180; break; case 2: angle2 = angle2; break; case 3: angle2 += 180; break; case 4: angle2 += 90; break; }} std::cout << "11111111" << std::endl; } /purple/ /yellow/ else if ( lenth[0]>lenth[1]&&lenth[0]>lenth[2]&&lenth[0]>lenth[5]&&lenth[0]>lenth[3]&&lenth[0]>lenth[4]) { //center2 = (approx[4]+approx[5])/2; //center = (center2+approx[1])/2; center = (approx[4]+approx[1])/2; if(width < height){ switch (start_number) { case 0: angle2 += 180; break; case 1: angle2 = angle2; break; case 2: angle2 += 180; break; case 3: angle2 = angle2; break; case 4: angle2 += 90; break; } } else{ switch (start_number) { case 0: angle2 -= 90; break; case 1: angle2 += 90; break; case 2: angle2 -= 90; break; case 3: angle2 += 90; break; case 4: angle2 += 180; break; } } std::cout << "000000" << std::endl; } /yellow*/ else if ( lenth[4]>lenth[1]&&lenth[4]>lenth[0]&&lenth[4]>lenth[5]&&lenth[4]>lenth[3]&&lenth[4]>lenth[2])/////yellow { //center2 = (approx[2]+approx[3])/2; // center = (center2+approx[5])/2; center = (approx[2]+approx[5])/2; if(width < height){ if(approx[5].y<approx[0].y&&approx[5].y<approx[1].y&&approx[5].y<approx[2].y&&approx[5].y<approx[3].y&&approx[5].y<approx[4].y){ switch (start_number) { case 0: angle2 = angle2; break; case 1: angle2 += 180; break; case 2: angle2 = angle2; break; case 3: angle2 += 180; break; case 4: angle2 -= 90; break; } } else { switch (start_number) { case 0: angle2 += 180; break; case 1: angle2 = angle2; break; case 2: angle2 += 180; break; case 3: angle2 = angle2; break; case 4: angle2 += 90; break; } } } else if(width > height){ switch (start_number) { case 0: angle2 += 90; break; case 1: angle2 -= 90; break; case 2: angle2 += 90; break; case 3: angle2 -= 90; break; case 4: angle2 = angle2; break; } } std::cout << "444444" << std::endl; } /yellow*********************/ } if(center.y<60) { center.y=60; } if(center.x<82) { center.x=82; } if(center.y>=60&¢er.x>=82) { //cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x1.2-92, center.y1.285-76.5,1); cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x, center.y,1); // std::cout << pixelPointMat << std::endl; // pixelPointMat = (cv::Mat_<double>(3, 1) << 640, 360, 1); cameraPointMat = objectHeight * K.inv() * pixelPointMat; // std::cout << "相机坐标系下的三维坐标：" << std::endl; // std::cout << cameraPointMat << std::endl; obj_camera_frame1.setX(cameraPointMat.at<double>(0,0)); obj_camera_frame1.setY(cameraPointMat.at<double>(1,0)); obj_camera_frame1.setZ(cameraPointMat.at<double>(2,0)); obj_robot_frame = camera_to_robot_ * obj_camera_frame1; std::cout<< " X :" << obj_robot_frame.getX() << std::endl; std::cout<< " Y :" << obj_robot_frame.getY() << std::endl; std::cout<< " Z :" << obj_robot_frame.getZ() << std::endl; if(angle2>176) angle2=176; if(angle2<-176) angle2=-176; std::cout << " angle2 :" << angle2 << std::endl; Grasping(obj_robot_frame.getX(),obj_robot_frame.getY(),obj_robot_frame.getZ(),angle2); } } else if(contour.size()>0&&contour.size()<280) { std::cout << "没有黄紫像素坐标：" << std::endl; }} } //绿色和蓝色 void Camera_TO_Robot_Process_GB(const std::vector<std::vector<cv::Point>>& contours, int start_number, int end_number) { std::cout << "green or blue" << std::endl; // 输出提示信息 cv::Point2f center; for(start_number; start_number < end_number; start_number++) { const std::vector<cv::Point>& contour = contours[start_number]; if (contour.size()>300) { // 检查轮廓是否为空 // 获取最小外接矩形 cv::RotatedRect minRect = cv::minAreaRect(contours[start_number]); center = minRect.center; double angleInRadians = minRect.angle * M_PI / 180.0; if(center.y<60) { center.y=60; } if(center.x<82) { center.x=82; } if(center.y>=60&¢er.x>=82) { cv::Mat pixelPointMat = (cv::Mat_<double>(3, 1) << center.x, center.y,1); std::cout << pixelPointMat << std::endl; cameraPointMat = objectHeight * K.inv() * pixelPointMat; //平面下的坐标转为相机三维坐标 //相机坐标转为基座标 obj_camera_frame1.setX(cameraPointMat.at<double>(0, 0)); obj_camera_frame1.setY(cameraPointMat.at<double>(1, 0)); obj_camera_frame1.setZ(cameraPointMat.at<double>(2, 0)); obj_robot_frame= camera_to_robot_ * obj_camera_frame1; std::cout << "坐标为: " << obj_robot_frame << std::endl; /****** 获取矩形的角度*/ double angle = minRect.angle; cv::Size2f size = minRect.size; double width = size.width; double height = size.height; //放平角度 if (width > height) { angle += 90; } Grasping(obj_robot_frame.getX(),obj_robot_frame.getY(),obj_robot_frame.getZ(),angle); } } else if(contour.size()>0&&contour.size()<100) { std::cout << "没有蓝绿像素坐标：" << std::endl; } } } }; /调用服务运行机械臂*************/ int move_lineb_test(xarm_msgs::Move srv, ros::ServiceClient client, float x_mm0, float y_mm0, float z_mm0, double roll0, double pitch0, double yaw0, float x_mm1, float y_mm1, float z_mm1, double roll1, double pitch1, double yaw1, float x_mm2, float y_mm2, float z_mm2, double roll2, double pitch2, double yaw2, float x_mm3, float y_mm3, float z_mm3, double roll3, double pitch3, double yaw3, float x_mm4, float y_mm4, float z_mm4, double roll4, double pitch4, double yaw4) { // 设置机械臂的运动速度 srv.request.mvvelo = 160; // 设置机械臂的运动加速度 srv.request.mvacc = 1000; // 设置机械臂的运动时间 srv.request.mvtime = 0; // 设置机械臂运动路径的圆角半径 srv.request.mvradii = 20; ROS_INFO("ZUOBIAOR: %f, %f,%f", x_mm1, y_mm1, z_mm1); std::vector<float> pose[5] = { {x_mm0, y_mm0, z_mm0, static_cast<float>(roll0), static_cast<float>(pitch0), static_cast<float>(yaw0)}, {x_mm1, y_mm1, z_mm1, static_cast<float>(roll1), static_cast<float>(pitch1), static_cast<float>(yaw1)}, {x_mm2, y_mm2, z_mm2, static_cast<float>(roll2), static_cast<float>(pitch2), static_cast<float>(yaw2)}, {x_mm3, y_mm3, z_mm3, static_cast<float>(roll3), static_cast<float>(pitch3), static_cast<float>(yaw3)}, {x_mm4, y_mm4, z_mm4, static_cast<float>(roll4), static_cast<float>(pitch4), static_cast<float>(yaw4)} }; for(int i = 0; i < 5; i++) { srv.request.pose = pose[i]; if(client.call(srv)) { ROS_INFO("%s\n", srv.response.message.c_str()); std::cout << "success111" << std::endl; } else { ROS_ERROR("Failed to call service move_lineb"); } } return 0; } bool is_at_pose(const geometry_msgs::Pose& current_pose, const geometry_msgs::Pose& target_pose, double tolerance = 0.005) { return std::abs(current_pose.position.x - target_pose.position.x) < tolerance && std::abs(current_pose.position.y - target_pose.position.y) < tolerance && std::abs(current_pose.position.z - target_pose.position.z) < tolerance; } //阻塞服务，检测是否到达目标位置 void control_suction_during_move(float x_mm4, float y_mm4, float z_mm4) { // 定义目标位 moveit::planning_interface::MoveGroupInterface arm("xarm7"); ros::AsyncSpinner spinner(1); spinner.start(); geometry_msgs::Pose target_pose4; target_pose4.position.x = x_mm4 / 1000.0; target_pose4.position.y = y_mm4 / 1000.0; target_pose4.position.z = (z_mm4-76.7398-13.9+8.5) / 1000.0; bool suction_off_triggered = false; // 循环检查当前位姿 ros::Rate rate(10); // 10 Hz 检查频率 while (ros::ok()) { geometry_msgs::Pose current_pose = arm.getCurrentPose().pose; // ROS_INFO("ZUOBIAO_SHISHI: %f, %f, %f", current_pose.position.x, current_pose.position.y, current_pose.position.z); // 检查是否到达关闭吸盘的位置 if (!suction_off_triggered && is_at_pose(current_pose, target_pose4)) { ROS_INFO("Reached suction off position, trying to turn off suction."); sleep(0.8); writeSpeed(0); writeSpeed(0); writeSpeed(0); sleep(0.5); ROS_INFO("Suction off commands sent."); suction_off_triggered = true; } // 如果两个条件都满足，退出循环 if (suction_off_triggered) { break; } rate.sleep(); } } int main(int argc, char argv) { // 初始化ROS节点 ros::init(argc, argv, "xarm_api"); ros::NodeHandle nh; XArmAPItest ic; nh.setParam("/xarm/wait_for_finish", true); ros::Publisher sleep_pub_ = nh.advertise<std_msgs::Float32>("/xarm/sleep_sec", 1); ros::ServiceClient motion_ctrl_client_ = nh.serviceClient<xarm_msgs::SetAxis>("/xarm/motion_ctrl"); ros::ServiceClient set_mode_client_ = nh.serviceClient<xarm_msgs::SetInt16>("/xarm/set_mode"); ros::ServiceClient set_state_client_ = nh.serviceClient<xarm_msgs::SetInt16>("/xarm/set_state"); ros::ServiceClient move_lineb_client_ = nh.serviceClient<xarm_msgs::Move>("/xarm/move_lineb"); xarm_msgs::SetAxis set_axis_srv_; xarm_msgs::SetInt16 set_int16_srv_; xarm_msgs::Move move_srv_; float x_mm0, y_mm0, z_mm0; double roll0, pitch0, yaw0; float x_mm1, y_mm1, z_mm1; double roll1, pitch1, yaw1; float x_mm2, y_mm2, z_mm2; double roll2, pitch2, yaw2; float x_mm3, y_mm3, z_mm3; double roll3, pitch3, yaw3; float x_mm4, y_mm4, z_mm4; double roll4, pitch4, yaw4; set_axis_srv_.request.id = 8; set_axis_srv_.request.data = 1; if(motion_ctrl_client_.call(set_axis_srv_)) { ROS_INFO("%s\n", set_axis_srv_.response.message.c_str()); } else { ROS_ERROR("Failed to call service motion_ctrl"); return 1; } set_int16_srv_.request.data = 0; if(set_mode_client_.call(set_int16_srv_)) { ROS_INFO("%s\n", set_int16_srv_.response.message.c_str()); } else { ROS_ERROR("Failed to call service set_mode"); return 1; } set_int16_srv_.request.data = 0; if(set_state_client_.call(set_int16_srv_)) { ROS_INFO("%s\n", set_int16_srv_.response.message.c_str()); } else { ROS_ERROR("Failed to call service set_state"); return 1; } nh.setParam("/xarm/wait_for_finish", false); std_msgs::Float32 sleep_msg; sleep_msg.data = 1.0; sleep_pub_.publish(sleep_msg); if(move_lineb_test(move_srv_, move_lineb_client_, x_mm0, y_mm0, z_mm0, roll0, pitch0, yaw0, x_mm1, y_mm1, z_mm1, roll1, pitch1, yaw1, x_mm2, y_mm2, z_mm2, roll2, pitch2, yaw2, x_mm3, y_mm3, z_mm3, roll3, pitch3, yaw3, x_mm4, y_mm4, z_mm4, roll4, pitch4, yaw4) == 1) return 1; // 调用 control_suction_during_move 函数 control_suction_during_move(x_mm4, y_mm4, z_mm4); sleep(0.8); nh.setParam("/xarm/wait_for_finish", true); while(ros::ok()) { ros::spinOnce(); } return 0; }解释一下

OpenCV 提供了 undistortPoints 函数用于去除像素坐标的畸变影响，并结合相机外参（旋转和平移向量）将其转换为世界坐标系下的三维坐标 [^1]。 cpp cv::Mat cameraMatrix = (cv::Mat_(3, 3) , 0, cx, 0, fy, ...

旋转矩阵计算欧拉角

由旋转矩阵计算绕X、Y、Z轴旋转角度的计算公式，直接代入公式求解即可，适用于3*3旋转矩阵的计算。

从旋转矩阵计算欧拉角代码

02.《大数据》配套之二：-数据采集与预处理PPT.ppt

PLC的编程实例电机正反转控制演示幻灯片.ppt

计算机组成原理所有概念.ppt

计算机配许课上C语言数据类型和学习框架思维导图

课上总结

yaochao_edu_source_crawler_2920_1754319247437.zip

计算机基础知识点总结PPT.ppt

877. 石子游戏（高频题）

877. 石子游戏题目解题思路代码题目亚历克斯和李用几堆石子在做游戏。偶数堆石子排成一行，每堆都有正整数颗石子 piles[i] 。游戏以谁手中的石子最多来决出胜负。石子的总数是奇数，所以没有平局。亚历克斯和李轮流进行，亚历克斯先开始。每回合，玩家从行的开始或结束处取走整堆石头。这种情况一直持续到没有更多的石子堆为止，此时手中石子最多的玩家获胜。假设亚历克斯和李都发挥出最佳水平，当亚历克斯赢得比赛时返回 true ，当李赢得比赛时返回 false 。解题思路 P415 代码 class S

网络虚拟货币对现实经济潜在影响的分析.doc

C++如何将欧拉角转为旋转矩阵

相关推荐

eigen 旋转矩阵 按照RxRyRz或者RzRyRx转欧拉角

欧拉角求旋转矩阵c++

python 和c++实现旋转矩阵到欧拉角的变换方式

Qt如何将欧拉角转为旋转矩阵

如何将欧拉角转为旋转矩阵float[3][3]

在手眼标定中，现在有一组数据为[-0.3969707, -0.460018, 0.3899877, 90.2261, -168.2015, 89.7748]，分别代表机械手的位置XYZ和旋转角RXRYRZ，旋转方向为xyz，如何将它转为旋转矩阵和平移向量，结合opencv和eigen请用C++实现

Qt中如何将旋转角度转为欧拉角

c++ opengl四元数旋转

c++使用Eigen3来实现线激光相机y,z值转换为机器人坐标

旋转矩阵计算欧拉角

从旋转矩阵计算欧拉角代码

02.《大数据》配套之二：-数据采集与预处理PPT.ppt

PLC的编程实例电机正反转控制演示幻灯片.ppt

计算机组成原理所有概念.ppt

计算机配许课上C语言数据类型和学习框架思维导图

推荐-办公软件学习.ppt

yaochao_edu_source_crawler_2920_1754319247437.zip

计算机基础知识点总结PPT.ppt

877. 石子游戏（高频题）

网络虚拟货币对现实经济潜在影响的分析.doc

大家在看

PL2303驱动ForWindows11.zip

无外部基准电压时STM32L151精确采集ADC电压

kb4474419和kb4490628系统补丁.rar

XposedDetector

超实用zimo21取字模软件.7z

最新推荐

Python根据欧拉角求旋转矩阵的实例

根据旋转矩阵求旋转欧拉角

机器人旋转矩阵与欧拉角转换公式

02.《大数据》配套之二：-数据采集与预处理PPT.ppt

19年国赛服务器答案深度解析：网络搭建与应用

【VS2010模块化秘籍】：提升项目管理效率的10个技巧

数据分析师发展前景

Elasticsearch及IK分词器安装包资源汇总

从零开始：Axure插件开发入门指南，构建自定义Chrome工具

代码错误，修改代码

eigen 旋转矩阵　按照RxRyRz或者RzRyRx转欧拉角