为什么加入平面优化之后，回到原点的轨迹还是存在Z方向的漂移？

[ChenHuang20]

[TouchDeeper]

你可以确认一下平面优化是否开启了

[ChenHuang20]

我直接使用的是原始的参数如下，请问还需要在哪里开启呢？

imu: 1
wheel: 1
only_initial_with_wheel: 0 #只利用wheel进行初始化，不加入因子图
plane: 1
num_of_cam: 2

[ChenHuang20]

我确认已经开启平面优化了

[TouchDeeper]

你可以尝试一下加大平面约束的权重

[ChenHuang20]

加大权重，是减小以下3个值吗？ roll_n: 0.005 pitch_n: 0.005 zpw_n: 0.025 在 2022-11-04 19:47:20，"庄庭达" ***@***.***> 写道： 你可以尝试一下加大平面约束的权重 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[ChenHuang20]

无论怎么调整平面约束的权重，都无法使得轨迹回到水平位置

[ChenHuang20]

我发现不管开不开平面约束，出来的轨迹都是差不多的。平面约束好像用处不大？

[TouchDeeper]

你说的这个现象跟我之前的实验结论不太一致，你用的是哪个配置文件呢

[ChenHuang20]

使用的是你提供的默认文件,就是readme里的那句。是否需要更改？ 在 2022-11-05 16:03:29，"庄庭达" ***@***.***> 写道： 你说的这个现象跟我之前的实验结论不太一致，你用的是哪个配置文件呢 — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[ChenHuang20]

使用这个realsense_stereo_imu_config_ridgeback.yaml:

%YAML:1.0

#common parameters
#support: 1 imu 1 cam; 1 imu 2 cam: 2 cam; 
imu: 1
wheel: 1
only_initial_with_wheel: 0 #只利用wheel进行初始化，不加入因子图
plane: 1
num_of_cam: 2

imu_topic: "/imu/data"
wheel_topic: "/ridgeback_velocity_controller/odom"   #"/ridgeback_velocity_controller/odom"， “/odometry/filtered”
#TODO check the distortion
image0_topic: "/camera/infra1/image_rect_raw"
image1_topic: "/camera/infra2/image_rect_raw"
output_path: "/home/td/slam/vins_fusion_ws/src/VINS-Fusion/output"

cam0_calib: "infra1.yaml"
cam1_calib: "infra2.yaml"
image_width: 640
image_height: 480
   

# Extrinsic parameter between IMU and Camera.
estimate_extrinsic: 1   # 0  Have an accurate extrinsic parameters. We will trust the following imu^R_cam, imu^T_cam, don't change it.
  # 1  Have an initial guess about extrinsic parameters. We will optimize around your initial guess.
# 2  Don't know anything about extrinsic parameters. You don't need to give R,T. We will try to calibrate it. Do some rotation movement at beginning.
#If you choose 0 or 1, you should write down the following matrix.

extrinsic_type: 3 # 0 ALL
                  # 1 Only translation
                  # 2 Only Rotation
                  # 3 no z
                  # 4 no rotation and no z

body_T_cam0: !!opencv-matrix
   rows: 4
   cols: 4
   dt: d
   data: [ 0.0,   0.0,   1.0,  0.041,
          -1.0,   0.0,   0.0,  0.307,
           0.0,  -1.0,   0.0,  0.544,
           0,     0,     0,    1 ]

body_T_cam1: !!opencv-matrix
   rows: 4
   cols: 4
   dt: d
   data: [ 0.0,   0.0,   1.0,  0.041,
           -1.0,   0.0,   0.0, 0.258,
           0.0,  -1.0,   0.0,  0.544,
           0,     0,     0,    1 ]


# Extrinsic parameter between IMU and Wheel.
estimate_wheel_extrinsic: 0   # 0  Have an accurate extrinsic parameters. We will trust the following imu^R_cam, imu^T_cam, don't change it.
# 1  Have an initial guess about extrinsic parameters. We will optimize around your initial guess.
# 2  Don't know anything about extrinsic parameters. You don't need to give R,T. We will try to calibrate it. Do some rotation movement at beginning.
#If you choose 0 or 1, you should write down the following matrix.

extrinsic_type_wheel: 3 # 0 ALL
                        # 1 Only translation
                        # 2 Only Rotation
                        # 3 no z
                        # 4 no rotation and no z

#wheel to body
body_T_wheel: !!opencv-matrix
  rows: 4
  cols: 4
  dt: d
  data: [1, 0, 0, -0.208,
         0, 1, 0, 0.290,
         0, 0, 1, -0.168,
         0, 0, 0, 1]


#plane noise
#mono:0.01 stereo:0.005
roll_n: 0.01
#mono:0.01  stereo:0.005
pitch_n: 0.01
#mono:0.05 stereo:0.025
zpw_n: 0.05


#Multiple thread support
multiple_thread: 1

#feature traker paprameters
max_cnt: 150            # max feature number in feature tracking
min_dist: 30            # min distance between two features 
freq: 10                # frequence (Hz) of publish tracking result. At least 10Hz for good estimation. If set 0, the frequence will be same as raw image 
F_threshold: 1.0        # ransac threshold (pixel)
show_track: 1           # publish tracking image as topic
flow_back: 1            # perform forward and backward optical flow to improve feature tracking accuracy

#optimization parameters
max_solver_time: 0.04  # max solver itration time (ms), to guarantee real time
max_num_iterations: 8   # max solver itrations, to guarantee real time
keyframe_parallax: 10.0 # keyframe selection threshold (pixel)

#imu parameters       The more accurate parameters you provide, the better performance
acc_n: 0.1         # accelerometer measurement noise standard deviation. #0.2   0.04
gyr_n: 0.05        # gyroscope measurement noise standard deviation.     #0.05  0.004
acc_w: 7.1765713730075628e-04         # accelerometer bias random work noise standard deviation.  #0.002
gyr_w: 4.0e-05       # gyroscope bias random work noise standard deviation.     #4.0e-5
g_norm: 9.805         # gravity magnitude

#wheel parameters
# rad/s mono:0.004 stereo:0.002
wheel_gyro_noise_sigma: 0.004
#　m/s mono:0.01  stereo:0.006
wheel_velocity_noise_sigma: 0.01

estimate_wheel_intrinsic: 0
# 0  Have an accurate intrinsic parameters. We will trust the following sx, sy, sw, don't change it.
# 1  Have an initial guess about intrinsic parameters. We will optimize around your initial guess.
# 2  TODO Don't know anything about intrinsic parameters. You don't need to give sx, sy, sw. We will try to calibrate it. Do some rotation movement at beginning.
#If you choose 0 or 1, you should write down the following sx, sy, sw.
# wheel intrinsic
sx: 1.0
sy: 1.0
sw: 1.0


#unsynchronization parameters
estimate_td: 0                      # online estimate time offset between camera and imu
td: 0.00                             # initial value of time offset. unit: s. readed image clock + td = real image clock (IMU clock)
#unsynchronization parameters
estimate_td_wheel: 0                      # online estimate time offset between camera and wheel
td_wheel: 0.0                             # initial value of time offset. unit: s. readed image clock + td = real image clock (IMU clock)
#loop closure parameters
load_previous_pose_graph: 0        # load and reuse previous pose graph; load from 'pose_graph_save_path'
pose_graph_save_path: "/home/td/slam/vins_fusion_ws/src/VINS-Fusion/output/pose_graph" # save and load path
save_image: 0                   # save image in pose graph for visualization prupose; you can close this function by setting 0 

[TouchDeeper]

@ChenHuang20 可以试下把平面约束关了，对比下结果

[ChenHuang20]

开和关基本没有区别

[TouchDeeper]

@ChenHuang20 可以对比下单目情况下的平面约束开关