ZED2i+Franka实现眼在手外标定（aruco_ros+easy_hand）

先根据我上篇文章Moveit控制真实机械臂franka-优快云博客实现moveit对franka的控制

Step1:下载编译功能包：

1、安装vision_visp

(咱也不知道什么用，看别的文章都说装，那就装！)

cd ~/catkin_ws/src 
git clone -b noetic-devel https://github.com/lagadic/vision_visp.git
cd ..
catkin_make

可以删除其中的visp_tracker,visp_auto_tracker

2、安装aruco_ros

cd ~/catkin_ws/src 
git clone -b noetic-devel https://github.com/pal-robotics/aruco_ros
cd ..
catkin_make

测试aruco_ros

aruco码生成：Online ArUco markers generator

注意！！！dictionary需要选orginal ArUco，ID和size记住

启动zed2i相机：

roslaunch zed_wrapper zed2i.launch

修改aruco_ros-launch-single.launch文件

<arg name="markerId"        default="06"/>
<arg name="markerSize"      default="0.018"/>    <!-- in m -->

该为前面记下的ID和Size。

然后在此launch文件下运行single.launch

aruco码上出现坐标和方框就是识别成功！

3、安装easy_handeye

cd ~/catkin_ws/src 
git clone https://github.com/IFL-CAMP/easy_handeye
cd ..
catkin_make

Step2.编写launch文件

编写两个launch文件，放在目录catkin_ws/src/easy_handeye/easy_handeye/launch下面

zed_panda_calibrate.launch:

<?xml version="1.0" ?>
<launch>
    <arg name="namespace_prefix" default="panda_eob_calib"/>
 
    <!-- (start your robot's MoveIt! stack, e.g. include its moveit_planning_execution.launch) -->
    <include file="/home/admin1/catkin_ws/src/fu_moveit/launch/franka_control.launch">
        <arg name="robot_ip" value="172.16.0.3"/> <!-- set your robot ip -->
        <arg name="load_gripper" value="false"/>
    </include>        
    
    <!--<arg name="marker_frame" default="aruco_marker_frame"/>-->
        <arg name="marker_size"  default="0.18" />
        <arg name="marker_id"  default="06" />
        <arg name="eye"             default="left"/>
        <arg name="ref_frame"       default="zed2i_left_camera_frame"/>  <!-- leave empty and the pose will be published wrt param parent_name -->
        <arg name="corner_refinement" default="LINES" /> <!-- NONE, HARRIS, LINES, SUBPIX --> 
   
    <!-- start ArUco -->
    <rosparam file="$(find fu_moveit)/config/kinematics.yaml" command="load" />
    <node name="aruco_tracker" pkg="aruco_ros" type="single">
        <remap from="/camera_info" to="/zed2i/zed_node/left/camera_info" />
        <remap from="/image" to="/zed2i/zed_node/left/image_rect_color" />
        <param name="image_is_rectified" value="true"/>
        <param name="marker_size"        value="0.18"/>
        <param name="marker_id"          value="06"/>
        <param name="reference_frame"    value="zed2i_left_camera_frame"/>
        <param name="camera_frame"       value="zed2i_left_camera_frame"/>
        <param name="marker_frame"       value="aruco_marker_frame" />
        <param name="corner_refinement"  value="$(arg corner_refinement)" />
    </node>
 
</launch>

 zed_panda_easy.launch:

<?xml version="1.0" ?>
<launch>
 
    <arg name="namespace_prefix" default="panda_eob_calib" /> 
    <include file="$(find easy_handeye)/launch/calibrate.launch">
        <arg name="eye_on_hand" value="false"/>
        <arg name="namespace_prefix" value="$(arg namespace_prefix)"/>
        <arg name="move_group" value="panda_arm" />
        <!--<arg name="move_group" value="panda_manipulator"  doc="the name of move_group for the automatic robot motion with MoveIt!" /-->
        <arg name="freehand_robot_movement" value="false"/>
 
 
        <!-- fill in the following parameters according to your robot's published tf frames -->
        <arg name="robot_base_frame" value="world"/>
        <arg name="robot_effector_frame" value="panda_link8"/>
 
        <!-- fill in the following parameters according to your tracking system's published tf frames -->
        <arg name="tracking_base_frame" value="zed2i_left_camera_frame"/>
        <arg name="tracking_marker_frame" value="aruco_marker_frame"/>
        <arg name="robot_velocity_scaling" value="0.5" />
        <arg name="robot_acceleration_scaling" value="0.2" />
    </include>
 
    <!-- (publish tf after the calibration) -->
    <!-- roslaunch easy_handeye publish.launch eye_on_hand:=false namespace_prefix:=$(arg namespace_prefix) -->
 
</launch>

Step3.开始标定！

打开终端，先后运行以下三个launch，必须等前一个launch运行完再开下一个

roslaunch zed_wrapper zed2i.launch
roslaunch easy_handeye zed_panda_calibrate.launch
roslaunch easy_handeye zed_panda_easy.launch

 出来两个rviz和以下两个窗口

必须添加图像信息才能开始标定！打开rqt_image_view或者在rviz中add-image或者rqt-plugins-visualization-image view

以下内容取自：在Ubuntu20.04下对Kinect V2+Franka panda进行眼在手外标定_franka 手眼标定-优快云博客

开始标定：

同时，操作另外两个窗口

点击Check starting pose 弹出Ready to start:click on next pose即说明可以开始采样；随后点击Take Sample进行采样，之后点击Next Pose再点击Plan,如图所示：

依次执行，采样十七次，点击Compute进行计算输出转换矩阵，如图所示：

--------------------------------------------------------------我是分割线---------------------------------------------------

按照上述步骤，我点plan之后一直报错，找了很多博客都没解决，最终通过手动拖动franka完成了采样(就是手动拖动franka到一个位置然后点击Take Sample)，虽然很愚蠢，但卓有成效！点击compute一样有结果！

完结，撒花！