Kalibr IMU+相机联合标定（含相机标定）

记录一下自己的学习过程，仅供参考

1 IMU标定

IMU标定过程 含ceres-solver code_utils imu_utils安装

2 相机标定

2.1 录包

帧率改为4 曝光能看清就行（0.5）

source devel/setup.bash
roslaunch mvsua_ros_driver camera_driver.launch	//驱动你自己的相机
rosbag record /mvsua_cam/image_raw1	//改为你自己相机的话题名

打开可视化界面

rqt_image_view

2.2 Kalibr标定相机

Kalibr官方wiki
1、创建checkerboard.yaml文件
这里是12*9 边长为3cm的棋盘格，一定要注意内角不是格子数

target_type: 'checkerboard' #gridtype
targetCols: 11              #number of internal chessboard corners 内角不是格子数 格子数为12*9
targetRows: 8               #number of internal chessboard corners
rowSpacingMeters: 0.03      #size of one chessboard square [m]
colSpacingMeters: 0.03      #size of one chessboard square [m]

2、运行Kalibr

rosrun kalibr kalibr_calibrate_cameras --bag cam.bag --topics /mvsua_cam/image_raw1 --models pinhole-equi --target checkerboard.yaml

models根据需要选择：

Kalibr supports the following projection models:

pinhole camera model (pinhole)1
(intrinsics vector: [fu fv pu pv])
omnidirectional camera model (omni)2
(intrinsics vector: [xi fu fv pu pv])
double sphere camera model (ds)3
(intrinsics vector: [xi alpha fu fv pu pv])
extended unified camera model (eucm)4
(intrinsics vector: [alpha beta fu fv pu pv])

Kalibr supports the following distortion models:

radial-tangential (radtan)*
(distortion_coeffs: [k1 k2 r1 r2])
equidistant (equi)**
(distortion_coeffs: [k1 k2 k3 k4])
fov (fov)5
(distortion_coeffs: [w])
none (none)
(distortion_coeffs: [])

运行完之后生成结果：重投影误差1以内比较好

3 联合标定

1、录包
同时录制imu和相机的topic
2、根据imu标定结果创建imu.yaml

#Accelerometers
accelerometer_noise_density: 4.84e-02   #Noise density (continuous-time)
accelerometer_random_walk:   4.67e-03   #Bias random walk

#Gyroscopes
gyroscope_noise_density:     1.44e-03   #Noise density (continuous-time)
gyroscope_random_walk:       6.26e-05   #Bias random walk

rostopic:                    /mobile_base/sensors/imu_data_origin      #the IMU ROS topic
update_rate:                 200.0      #Hz (for discretization of the values above)

3、cam.yaml可以使用上一步标定生成的结果
4、联合标定

rosrun kalibr kalibr_calibrate_imu_camera --bag imu_cam.bag --cam cam.yaml --imu imu.yaml --target checkerboard.yaml

4 标定结果

Normalized Residuals：

相机、IMU标定结果：