Skip to content
  • Categories
  • Recent
  • Tags
  • Popular
  • Users
  • Groups
Collapse
Brand Logo

ModalAI Forum

  1. ModalAI Support Forum
  2. Ask your questions right here!
  3. Timeline for open-vins documentation

Timeline for open-vins documentation

Scheduled Pinned Locked Moved Ask your questions right here!
28 Posts 9 Posters 8.3k Views 4 Watching
  • Oldest to Newest
  • Newest to Oldest
  • Most Votes
Reply
  • Reply as topic
Log in to reply
This topic has been deleted. Only users with topic management privileges can see it.
  • Viktor GalV Viktor Gal

    @Zachary-Lowell-0 so those params are after running voxl-configure-mpa; i still dont understand why there's a diff between the apparently same v2 starlings.

    i can run the imu_apps calibration and retry OV.

    Viktor GalV Offline
    Viktor GalV Offline
    Viktor Gal
    Contributor
    wrote on last edited by
    #17

    @Zachary-Lowell-0 hey so i did what you suggested, here's the full output of openvins server but it's the very same behaviour; the qvio is nicely picking up the movements, openvins is just stationary on the portal. here are the detailed logs:

    voxl2:~$ voxl-open-vins-server --verbosity 0
    Setting printing level to: SILENT
    Setting printing level to: ALL
    Loading our own config file
    =================================================================
    =========================AUTO RESET==============================
    en_auto_reset:                    1
    auto_reset_max_velocity:          20.000
    auto_reset_max_v_cov_instant:      0.100
    auto_reset_max_v_cov:              0.100
    auto_reset_max_v_cov_timeout_s:    0.500
    auto_reset_min_features:          1
    auto_reset_min_feature_timeout_s:  3.000
    auto_fallback_timeout_s:  3.000
    auto_fallback_min_v:  0.600
    en_cont_yaw_checks: false
    fast_yaw_thresh:  5.000
    fast_yaw_timeout_s:  1.750
    ===========================STATE=================================
    do fej:                           true
    imu avg:                          true
    use rk4 integration:              true
    cam to imu refinement:            true
    cam intrins refinement:           true
    cam imu ts refinement:            true
    max clone size:                   8
    max slam features:                40
    max slam in update:               25
    max msckf in update:              25
    feat rep msckf:                   ANCHORED_MSCKF_INVERSE_DEPTH
    feat rep slam:                    ANCHORED_MSCKF_INVERSE_DEPTH
    cam imu time offset:              0.00150
    slam delay:                       1.00000
    =================================================================
    =====================INERTIAL INITIALIZER========================
    gravity mag:                      9.80665
    init window time:                 1.00000
    init imu thresh:                  1.00000
    init dyn use:                         false
    =================================================================
    ==========================IMU NOISE==============================
    imu sigma w:                      0.00014
    imu sigma wb:                     0.00000
    imu sigma a:                      0.00389
    imu sigma ab:                     0.00006
    =================================================================
    ========================FEATURE NOISE============================
    msckf chi^2 multiplier:           1.00000
    msckf sigma px:                   1.00000
    slam chi^2 multiplier:            40.00000
    slam sigma px:                    1.80000
    zupt chi^2 multiplier:            1.00000
    zupt sigma px:                    1.00000
    =================================================================
    =============================ZUPT================================
    try zupt:                         true
    zupt max velocity:                0.03000
    zupt only at beginning:           true
    zupt noise multiplier:            1.00000
    zupt max disparity:               3.00000
    =================================================================
    use mask:                         true
    use stereo:                       false
    use baro:                       false
    =================================================================
    ========================FEATURE INITIALIZER======================
    triangulate 1d:                   false
    refine features:                  false
    max runs:                         5
    init lamda:                       0.00100
    max lamda:                        10000000000.00000
    min dx:                           0.00000
    min dcost:                        0.00000
    lam mult:                         10.00000
    min dist:                         0.10000
    max dist:                         60.00000
    max baseline:                     40.00000
    num_opencv_threads:                  9
    fast_threshold:                  15
    histogram_method:                  HISTOGRAM
    knn_ratio:                  0.70000
    track_frequency:                  15.00000
    publish_frequency:                  3
    imu takeoff accel thresh:                  0.10000
    force init on takeoff:                  false
    use takeoff_camera_as:                  0
    takeoff_threshold(m):                  -0.100000
    publish stats:                  false
    max_allowable_cep:                  1.00000
    force FLU to NED transform:                  false
    VIO always on (for bench testing):                  false
    =================================================================
    =================================================================
    Number of cameras found for this config: 1
    [INFO] External feature trackers being used? false
    camera count: 1
    MONO mode (using cam idx: 0)
    Number of cameras found for this config: 1
    [INFO] External feature trackers being used? false
    camera count: 1
    MONO mode (using cam idx: 0)
    Skipping camera group index: 2, not enabled
    Loading OpenCV intrinsics cal file from: /data/modalai/opencv_tracking_front_intrinsics.yml
    Loading OpenCV intrinsics cal file from: /data/modalai/opencv_tracking_down_intrinsics.yml
    cam_load_extrinsics_file
    World Rotation Matrix:
     1.000000  0  0
     0  1.000000  0
     0  0  1.000000
    
    load extrinsics 2
    Using extrinsics under camera ID: tracking_front
    [INFO] Camera: 0 -- converted extrinsics *in FLU* are: Roll: -0.000000, Pitch 90.000000, Yaw -90.000000
    (local) CAMERA ROTATION MATRIX:
    [0, 0, 1;
     -1, 0, 0;
     0, -1, 0]
    (local) CAMERA TRANSLATION:
      0.037
         -0
    -0.0006
    Using extrinsics under camera ID: tracking_down
    [INFO] Camera: 1 -- converted extrinsics *in FLU* are: Roll: -0.000000, Pitch 180.000000, Yaw -360.000000
    (local) CAMERA ROTATION MATRIX:
    [-1, 0, 0;
     0, 1, 0;
     0, 0, -1]
    (local) CAMERA TRANSLATION:
    -0.08825
      0.0045
    -0.00269
    Done transfer camera configuration transfer
    =================================================================
    ==========================CAMERA 0================================
    name:                             tracking_front
    mode:                             MONO
    camera matrix:
    460.27780 0.00000 621.21213
    0.00000 459.85438 408.51998
    0.00000 0.00000 1.00000
    
    distortion coefficients:
    0.05324 0.02329 -0.00926 0.00026
    ==========================CAMERA 1================================
    name:                             tracking_front
    mode:                             MONO
    camera matrix:
    462.04073 0.00000 660.98324
    0.00000 462.09738 397.45450
    0.00000 0.00000 1.00000
    
    distortion coefficients:
    0.04984 0.04004 -0.02646 0.00611
    =============================KLT=================================
    num_features_to_track:            30
    grid_x:                           5
    grid_y:                           5
    min_pix_dist:                     50
    pyramid_levels:                   5
    [block] window_size:                 (25 x 25)
    ===========================KLTGYRO===============================
    tmp_imu_name:                         imu_apps
    tracking_front wrt imu_apps:
    0.00000 0.00000 1.00000
    -1.00000 0.00000 0.00000
    0.00000 -1.00000 0.00000
    tracking_down wrt imu_apps:
    -1.00000 0.00000 0.00000
    0.00000 1.00000 0.00000
    0.00000 0.00000 -1.00000
    Loading external config file
    =====> imu to read from: imu
    OpenVINS using fisheye camera
    OpenVINS FINAL Cam extrinsics to IMU for tracking_front: 7
        -0.5
        0.5
       -0.5
        0.5
      0.037
         -0
    -0.0006
    OpenVINS using fisheye camera
    OpenVINS FINAL Cam extrinsics to IMU for tracking_down: 7
     -1.22465e-16
              -1
     -7.4988e-33
    -6.12323e-17
        -0.08825
          0.0045
        -0.00269
     ++++++++++++++++++++++++++
    
    OpenVINS reports number of cameras being used, regardless of delivery method (pipe, merge, etc...):  2
    
     ++++++++++++++++++++++++++
    VioManagerOptions.h:108 ESTIMATOR PARAMETERS:
    StateOptions.h:108   - use_fej: 1
    StateOptions.h:109   - use_imuavg: 1
    StateOptions.h:110   - use_rk4int: 1
    StateOptions.h:111   - calib_cam_extrinsics: 1
    StateOptions.h:112   - calib_cam_intrinsics: 1
    StateOptions.h:113   - calib_cam_timeoffset: 1
    StateOptions.h:114   - max_clones: 8
    StateOptions.h:115   - max_slam: 40
    StateOptions.h:116   - max_slam_in_update: 25
    StateOptions.h:117   - max_msckf_in_update: 25
    StateOptions.h:118   - max_aruco: 1024
    StateOptions.h:119   - max_cameras: 2
    StateOptions.h:120   - feat_rep_msckf: ANCHORED_MSCKF_INVERSE_DEPTH
    StateOptions.h:121   - feat_rep_slam: ANCHORED_MSCKF_INVERSE_DEPTH
    StateOptions.h:122   - feat_rep_aruco: GLOBAL_3D
    InertialInitializerOptions.h:125 INITIALIZATION SETTINGS:
    InertialInitializerOptions.h:150   - init_window_time: 1.00
    InertialInitializerOptions.h:151   - init_imu_thresh: 1.00
    InertialInitializerOptions.h:152   - init_max_disparity: 3.00
    InertialInitializerOptions.h:153   - init_max_features: 0.00
    InertialInitializerOptions.h:171   - init_dyn_use: 0
    InertialInitializerOptions.h:172   - init_dyn_mle_opt_calib: 0
    InertialInitializerOptions.h:173   - init_dyn_mle_max_iter: 20
    InertialInitializerOptions.h:174   - init_dyn_mle_max_threads: 20
    InertialInitializerOptions.h:175   - init_dyn_mle_max_time: 2.00
    InertialInitializerOptions.h:176   - init_dyn_num_pose: 8
    InertialInitializerOptions.h:177   - init_dyn_min_deg: 45.00
    InertialInitializerOptions.h:178   - init_dyn_inflation_ori: 1.00e+01
    InertialInitializerOptions.h:179   - init_dyn_inflation_vel: 1.00e+01
    InertialInitializerOptions.h:180   - init_dyn_inflation_bg: 1.00e+02
    InertialInitializerOptions.h:181   - init_dyn_inflation_ba: 1.00e+02
    InertialInitializerOptions.h:182   - init_dyn_min_rec_cond: 1.00e-15
    InertialInitializerOptions.h:188   - init_dyn_bias_g: 0.00, 0.00, 0.00
    InertialInitializerOptions.h:189   - init_dyn_bias_a: 0.00, 0.00, 0.00
    InertialInitializerOptions.h:216 NOISE PARAMETERS:
    InertialInitializerOptions.h:224   - gyroscope_noise_density: 0.000170
    InertialInitializerOptions.h:225   - accelerometer_noise_density: 0.00200
    InertialInitializerOptions.h:226   - gyroscope_random_walk: 0.0000194
    InertialInitializerOptions.h:227   - accelerometer_random_walk: 0.003000
    InertialInitializerOptions.h:228   - sigma_pix: 1.00
    InertialInitializerOptions.h:318 STATE PARAMETERS:
    InertialInitializerOptions.h:319   - gravity_mag: 9.8066
    InertialInitializerOptions.h:320   - gravity: 0.000, 0.000, 9.807
    InertialInitializerOptions.h:321   - num_cameras: 2
    InertialInitializerOptions.h:322   - use_stereo: 0
    InertialInitializerOptions.h:323   - downsize cameras: 0
    InertialInitializerOptions.h:332   - calib_camimu_dt: 0.0000
    InertialInitializerOptions.h:348 cam_0_fisheye:1
    cam_0_wh:
    1280 x 800
    cam_0_intrinsic(0:3):
    460.278 459.854 621.212  408.52
    cam_0_intrinsic(4:7):
      0.0532387   0.0232891  -0.0092603 0.000263136
    cam_0_extrinsic(0:3):
    -0.5  0.5 -0.5  0.5
    cam_0_extrinsic(4:6):
      0.037      -0 -0.0006
    T_C0toI:
               0  2.22045e-16            1       0.0006
              -1  2.22045e-16            0        0.037
    -2.22045e-16           -1  2.22045e-16  8.34888e-18
               0            0            0            1
    
    InertialInitializerOptions.h:348 cam_1_fisheye:1
    cam_1_wh:
    1280 x 800
    cam_1_intrinsic(0:3):
    462.041 462.097 660.983 397.455
    cam_1_intrinsic(4:7):
     0.0498391   0.040037 -0.0264578  0.0061076
    cam_1_extrinsic(0:3):
    -1.22465e-16           -1  -7.4988e-33 -6.12323e-17
    cam_1_extrinsic(4:6):
    -0.08825   0.0045 -0.00269
    T_C1toI:
              -1  2.44929e-16  1.22465e-16     -0.08825
     2.44929e-16            1            0      -0.0045
    -1.22465e-16  2.99952e-32           -1     -0.00269
               0            0            0            1
    
    VioManagerOptions.h:121   - dt_slam_delay: 1.0
    VioManagerOptions.h:122   - zero_velocity_update: 1
    VioManagerOptions.h:123   - zupt_max_velocity: 0.03
    VioManagerOptions.h:124   - zupt_noise_multiplier: 1.00
    VioManagerOptions.h:125   - zupt_max_disparity: 3.0000
    VioManagerOptions.h:126   - zupt_only_at_beginning?: 1
    VioManagerOptions.h:127   - record timing?: 0
    VioManagerOptions.h:128   - record timing filepath: ov_msckf_timing.txt
    VioManagerOptions.h:155 NOISE PARAMETERS:
    NoiseManager.h:62   - gyroscope_noise_density: 0.000140
    NoiseManager.h:63   - accelerometer_noise_density: 0.00389
    NoiseManager.h:64   - gyroscope_random_walk: 0.0000004
    NoiseManager.h:65   - accelerometer_random_walk: 0.000055
    VioManagerOptions.h:179   Updater MSCKF Feats:
    UpdaterOptions.h:45     - chi2_multipler: 1.0
    UpdaterOptions.h:46     - sigma_pix: 1.00
    VioManagerOptions.h:181   Updater SLAM Feats:
    UpdaterOptions.h:45     - chi2_multipler: 40.0
    UpdaterOptions.h:46     - sigma_pix: 1.80
    VioManagerOptions.h:183   Updater ARUCO Tags:
    UpdaterOptions.h:45     - chi2_multipler: 5.0
    UpdaterOptions.h:46     - sigma_pix: 1.00
    VioManagerOptions.h:185   Updater ZUPT:
    UpdaterOptions.h:45     - chi2_multipler: 1.0
    UpdaterOptions.h:46     - sigma_pix: 1.00
    VioManagerOptions.h:292 STATE PARAMETERS:
    VioManagerOptions.h:293   - gravity_mag: 9.8100
    VioManagerOptions.h:294   - gravity: 0.000, 0.000, 9.810
    VioManagerOptions.h:295   - camera masks?: 1
    VioManagerOptions.h:304   - calib_camimu_dt: 0.0015
    VioManagerOptions.h:320 cam_0_fisheye:1
    cam_0_wh:
    1280 x 800
    cam_0_intrinsic(0:3):
    460.278 459.854 621.212  408.52
    cam_0_intrinsic(4:7):
      0.0532387   0.0232891  -0.0092603 0.000263136
    cam_0_extrinsic(0:3):
    -0.5  0.5 -0.5  0.5
    cam_0_extrinsic(4:6):
      0.037      -0 -0.0006
    T_C0toI:
               0  2.22045e-16            1       0.0006
              -1  2.22045e-16            0        0.037
    -2.22045e-16           -1  2.22045e-16  8.34888e-18
               0            0            0            1
    
    VioManagerOptions.h:320 cam_1_fisheye:1
    cam_1_wh:
    1280 x 800
    cam_1_intrinsic(0:3):
    462.041 462.097 660.983 397.455
    cam_1_intrinsic(4:7):
     0.0498391   0.040037 -0.0264578  0.0061076
    cam_1_extrinsic(0:3):
    -1.22465e-16           -1  -7.4988e-33 -6.12323e-17
    cam_1_extrinsic(4:6):
    -0.08825   0.0045 -0.00269
    T_C1toI:
              -1  2.44929e-16  1.22465e-16     -0.08825
     2.44929e-16            1            0      -0.0045
    -1.22465e-16  2.99952e-32           -1     -0.00269
               0            0            0            1
    
    VioManagerOptions.h:426 FEATURE TRACKING PARAMETERS:
    VioManagerOptions.h:427   - use_stereo: 0
    VioManagerOptions.h:428   - use_klt: 1
    VioManagerOptions.h:429   - use_aruco: 0
    VioManagerOptions.h:430   - downsize aruco: 1
    VioManagerOptions.h:431   - downsize cameras: 0
    VioManagerOptions.h:432   - num opencv threads: 9
    VioManagerOptions.h:433   - use multi-threading pubs: 1
    VioManagerOptions.h:434   - use multi-threading subs: 1
    VioManagerOptions.h:435   - num_pts: 30
    VioManagerOptions.h:436   - fast threshold: 15
    VioManagerOptions.h:437   - grid X by Y: 5 by 5
    VioManagerOptions.h:438   - min px dist: 50
    VioManagerOptions.h:439   - hist method: 1
    VioManagerOptions.h:440   - knn ratio: 0.700
    VioManagerOptions.h:441   - track frequency: 15.0
    VioManagerOptions.h:442   - limit imu propagation (windows based on init_imu+cam-imu delay): 1
    FeatureInitializerOptions.h:87 	- triangulate_1d: 0
    FeatureInitializerOptions.h:88 	- refine_features: 0
    FeatureInitializerOptions.h:89 	- max_runs: 5
    FeatureInitializerOptions.h:90 	- init_lamda: 0.001
    FeatureInitializerOptions.h:91 	- max_lamda: 10000000000.000
    FeatureInitializerOptions.h:92 	- min_dx: 0.0000010
    FeatureInitializerOptions.h:93 	- min_dcost: 0.0000010
    FeatureInitializerOptions.h:94 	- lam_mult: 10.000
    FeatureInitializerOptions.h:95 	- min_dist: 0.100
    FeatureInitializerOptions.h:96 	- max_dist: 60.000
    FeatureInitializerOptions.h:97 	- max_baseline: 40.000
    FeatureInitializerOptions.h:98 	- max_cond_number: 600000.000
    Set camera intrinsics and extrinsics
    
    ====> Using Internal KLT feature tracker <====
    >>>>>>>>>> Current reference counts:
    trackDB: 1
    trackFEATS: 1
    propagator: 2
    updaterMSCKF: 1
    updaterSLAM: 1
    active_tracks_initializer: 1
    initializer: 1
    setting scheduler
    INFO: set FIFO priority successfully!
    create_server_pipes
    connecting client pipes
    Number of Cameras active: 2
    Camera merge --- > ch: 1 to cam id: 0
    Opening camera pipe: tracking_front
    tmp_camera_pipe_names.push_back(): tracking_front
    Camera merge --- > ch: 2 to cam id: 1
    Opening camera pipe: tracking_down
    tmp_camera_pipe_names.push_back(): tracking_down
    imu pipe name: imu_apps
    Camera processing thread is now locked to the following cores: 7
    VioManager.cpp:454 [TIME]: 0.0358 seconds for tracking
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0000 seconds
    VioManager.cpp:454 [TIME]: 0.0139 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,0 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0002 seconds
    VioManager.cpp:454 [TIME]: 0.0170 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,0 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0003 seconds
    VioManager.cpp:454 [TIME]: 0.0162 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,62 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0001 seconds
    VioManager.cpp:454 [TIME]: 0.0112 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,62 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0003 seconds
    VioManager.cpp:454 [TIME]: 0.0175 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,62 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0002 seconds
    VioManager.cpp:454 [TIME]: 0.0156 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,62 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0003 seconds
    VioManager.cpp:454 [TIME]: 0.0105 seconds for tracking
    InertialInitializer.cpp:127 [init]: not enough feats to compute disp: 0,62 < 15
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0003 seconds
    VioManager.cpp:454 [TIME]: 0.0169 seconds for tracking
    InertialInitializer.cpp:134 [init]: *** disparity is 0.000,0.000 (3.00 thresh)
    InertialInitializer.cpp:145 [init]: USING STATIC INITIALIZER METHOD!
    IMU samples too small, make your window smaller (params.init_window_time), can't initalize 241.141469 - 240.222589  (0.918880)
    StaticInitializer.cpp:55 [init-s]: unable to select window of IMU readings, not enough readings
    VioManagerHelper.cpp:174 [init]: failed initialization in 0.0004 seconds
    VioManager.cpp:454 [TIME]: 0.0115 seconds for tracking
    InertialInitializer.cpp:134 [init]: *** disparity is 0.000,0.000 (3.00 thresh)
    InertialInitializer.cpp:145 [init]: USING STATIC INITIALIZER METHOD!
    StaticInitializer.cpp:104 [init-s]: IMU excitation stats: 0.029,0.014
    VioManagerHelper.cpp:138 [init]: successful initialization in 0.0008 seconds
    VioManagerHelper.cpp:140 [init]: orientation = -0.0029, 0.0025, 0.0000, 1.0000
    VioManagerHelper.cpp:142 [init]: bias gyro = 0.0025, 0.0000, 0.0016
    VioManagerHelper.cpp:144 [init]: velocity = 0.0000, 0.0000, 0.0000
    VioManagerHelper.cpp:146 [init]: bias accel = 0.0000, 0.0000, -0.0046
    VioManagerHelper.cpp:148 [init]: position = 0.0000, 0.0000, 0.0000
    VioManagerHelper.cpp:168 [init]: moved the state forward 0.00 seconds
    StateHelper.cpp:686 new augment clone at timestamp: 241.367844
    waiting for enough clone states (1 of 5)....
    [INFO] TAKEOFF: DISPARITY disturbed, possible motion detected. Avg: 91.364071 (3.000000)	# feats: 5 (20).
    UpdaterZeroVelocity.cpp:242 [ZUPT]: failed disparity (91.364 > 3.000, 5 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.003 (chi2 99.575 < 113.145)
    [INFO] TAKEOFF: DISPARITY disturbed, possible motion detected. Avg: 0.000000 (3.000000)	# feats: 5 (20).
    UpdaterZeroVelocity.cpp:242 [ZUPT]: failed disparity (0.000 > 3.000, 5 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.007 (chi2 62.236 < 106.395)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.004 (chi2 88.421 < 119.871)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.004 (chi2 96.190 < 119.871)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.003 (chi2 132.910 < 119.871)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.003 (chi2 95.202 < 119.871)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.003 (chi2 76.613 < 106.395)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.002 (chi2 105.215 < 113.145)
    UpdaterZeroVelocity.cpp:233 [ZUPT]: passed disparity (0.000 < 3.000, 40 features)
    UpdaterZeroVelocity.cpp:255 [ZUPT]: accepted |v_IinG| = 0.002 (chi2 86.990 < 119.871)
    
    1 Reply Last reply
    0
    • Zachary Lowell 0Z Offline
      Zachary Lowell 0Z Offline
      Zachary Lowell 0
      ModalAI Team
      wrote on last edited by
      #18

      @Viktor-Gal said in Timeline for open-vins documentation:

      i can run the imu_apps calibration and retry OV.

      1 Reply Last reply about 15 hours ago Reply

      a day ago

      @Zachary-Lowell-0 so those params are after running voxl-configure-mpa; i still dont understand why there's a diff between the apparently same v2 starlings.

      i can run the imu_apps calibration and retry OV.

      RE: No you are sending the PX4 parameters not the voxl2 parameters. Runnign voxl-configure-mpa will configure some px4 parameters, but after you open up qgc and then calibrateall the sensors in there, it will change some of those parameters, hence why those specific to imu calibration and mag cal are different from drone to drone. I highly recommend sticking to qvio with single camera tracking for now as OVINS is still heavily in development.

      Viktor GalV 1 Reply Last reply
      0
      • Zachary Lowell 0Z Zachary Lowell 0

        @Viktor-Gal said in Timeline for open-vins documentation:

        i can run the imu_apps calibration and retry OV.

        1 Reply Last reply about 15 hours ago Reply

        a day ago

        @Zachary-Lowell-0 so those params are after running voxl-configure-mpa; i still dont understand why there's a diff between the apparently same v2 starlings.

        i can run the imu_apps calibration and retry OV.

        RE: No you are sending the PX4 parameters not the voxl2 parameters. Runnign voxl-configure-mpa will configure some px4 parameters, but after you open up qgc and then calibrateall the sensors in there, it will change some of those parameters, hence why those specific to imu calibration and mag cal are different from drone to drone. I highly recommend sticking to qvio with single camera tracking for now as OVINS is still heavily in development.

        Viktor GalV Offline
        Viktor GalV Offline
        Viktor Gal
        Contributor
        wrote on last edited by
        #19

        @Zachary-Lowell-0 i hear you, but could you maybe give some pointers how one could make the qvio more stable? as i said it is at the moment only has 25% success rate: meaning 1 out of 5 flights i dont have problems with the qvio, the rest of the time the drone becomes unstable and crashes.

        ModeratorM 1 Reply Last reply
        0
        • Viktor GalV Viktor Gal

          @Zachary-Lowell-0 i hear you, but could you maybe give some pointers how one could make the qvio more stable? as i said it is at the moment only has 25% success rate: meaning 1 out of 5 flights i dont have problems with the qvio, the rest of the time the drone becomes unstable and crashes.

          ModeratorM Offline
          ModeratorM Offline
          Moderator
          ModalAI Team
          wrote on last edited by
          #20

          @Viktor-Gal have you read through the troubleshooting VIO documentation? https://docs.modalai.com/flying-with-vio/#troubleshooting-vio

          Viktor GalV 1 Reply Last reply
          0
          • ModeratorM Moderator

            @Viktor-Gal have you read through the troubleshooting VIO documentation? https://docs.modalai.com/flying-with-vio/#troubleshooting-vio

            Viktor GalV Offline
            Viktor GalV Offline
            Viktor Gal
            Contributor
            wrote on last edited by
            #21

            @Moderator yep, back and forth. unfortunately no change.

            Viktor GalV 1 Reply Last reply
            0
            • Viktor GalV Viktor Gal

              @Moderator yep, back and forth. unfortunately no change.

              Viktor GalV Offline
              Viktor GalV Offline
              Viktor Gal
              Contributor
              wrote on last edited by
              #22

              @Moderator et.al. i just want to follow up that switching to the nightly builds that uses out of box the openvins implementation for VIO, is significantly better than the qvio. in other words, it actually fixed all the problems that we were facing with the QVIO.

              Eric KatzfeyE dleeD 2 Replies Last reply
              1
              • Viktor GalV Viktor Gal

                @Moderator et.al. i just want to follow up that switching to the nightly builds that uses out of box the openvins implementation for VIO, is significantly better than the qvio. in other words, it actually fixed all the problems that we were facing with the QVIO.

                Eric KatzfeyE Offline
                Eric KatzfeyE Offline
                Eric Katzfey
                ModalAI Team
                wrote on last edited by
                #23

                @Viktor-Gal Excellent, thanks for the feedback!

                1 Reply Last reply
                0
                • S Offline
                  S Offline
                  star123
                  wrote on last edited by star123
                  #24

                  I'll put a question here since it's open-vins documentation related. I also am successfully running nightly builds of open-vins, and it is working well for me.

                  I would like to add image masks to the tracking cameras to block out certain areas. I see the option "use_mask" in voxl-open-vins-server.conf, which I can set to true. I can't figure out where to put the mask files, or how to specify their path and file name?

                  If this should be a separate thread, please let me know and I can move my question elsewhere, cheers!

                  dleeD Cliff WongC 2 Replies Last reply
                  0
                  • A Aaky referenced this topic on
                  • S star123

                    I'll put a question here since it's open-vins documentation related. I also am successfully running nightly builds of open-vins, and it is working well for me.

                    I would like to add image masks to the tracking cameras to block out certain areas. I see the option "use_mask" in voxl-open-vins-server.conf, which I can set to true. I can't figure out where to put the mask files, or how to specify their path and file name?

                    If this should be a separate thread, please let me know and I can move my question elsewhere, cheers!

                    dleeD Offline
                    dleeD Offline
                    dlee
                    Regular
                    wrote on last edited by
                    #25

                    @star123 What does nightly mean? The latest version of voxl-open-vins-server or voxl-suite. Thanks.

                    tomT 1 Reply Last reply
                    0
                    • dleeD dlee

                      @star123 What does nightly mean? The latest version of voxl-open-vins-server or voxl-suite. Thanks.

                      tomT Offline
                      tomT Offline
                      tom
                      admin
                      wrote on last edited by
                      #26

                      @dlee We publish "nightly" SDK builds at https://storage.googleapis.com/platform-nightlies/voxl2/voxl2_SDK_nightly_20250213.tar.gz (URL will match yesterday's date, PST)

                      This takes the latest of each repo's dev branch as well as the latest system image built every night

                      1 Reply Last reply
                      0
                      • S star123

                        I'll put a question here since it's open-vins documentation related. I also am successfully running nightly builds of open-vins, and it is working well for me.

                        I would like to add image masks to the tracking cameras to block out certain areas. I see the option "use_mask" in voxl-open-vins-server.conf, which I can set to true. I can't figure out where to put the mask files, or how to specify their path and file name?

                        If this should be a separate thread, please let me know and I can move my question elsewhere, cheers!

                        Cliff WongC Offline
                        Cliff WongC Offline
                        Cliff Wong
                        ModalAI Team
                        wrote on last edited by Cliff Wong
                        #27

                        @star123 At this time masks are supported through programmatic means only.

                        I'll add it to our feature list for the next release and should be in our nightly builds by mid-next week. If you take a look at /data/modalai/zero_ref.jpg is an example format to give you an idea when the feature is ready.

                        1 Reply Last reply
                        0
                        • Viktor GalV Viktor Gal

                          @Moderator et.al. i just want to follow up that switching to the nightly builds that uses out of box the openvins implementation for VIO, is significantly better than the qvio. in other words, it actually fixed all the problems that we were facing with the QVIO.

                          dleeD Offline
                          dleeD Offline
                          dlee
                          Regular
                          wrote on last edited by
                          #28

                          @Viktor-Gal I am troubling localization with QVIO. So I wanted to change vio algorithm to OpenVINS. I flashed my drone to nightly version but openvins still very unstable. Could you give me some guidance for openvins?

                          Questions

                          1. openvins configuration & versions
                          2. Did you enable the voxl-feature-tracker?
                          3. camera settings (single or multi, camera model, etc)

                          My status

                          • voxl-suite: 1.4.1-202502132201
                          • Camera: old-version of tracking camera (640x480) - link
                          • voxl-open-vins-server: 0.3.0-202502061726
                          • voxl-feature-tracker: 0.5.2-202412171357
                          1 Reply Last reply
                          0

                          Hello! It looks like you're interested in this conversation, but you don't have an account yet.

                          Getting fed up of having to scroll through the same posts each visit? When you register for an account, you'll always come back to exactly where you were before, and choose to be notified of new replies (either via email, or push notification). You'll also be able to save bookmarks and upvote posts to show your appreciation to other community members.

                          With your input, this post could be even better 💗

                          Register Login
                          Reply
                          • Reply as topic
                          Log in to reply
                          • Oldest to Newest
                          • Newest to Oldest
                          • Most Votes


                          ModalAI
                          Categories Recent Tags ModalAI.com Docs
                          © 2026 ModalAI® · Accelerating autonomy for smaller, smarter, safer drones · Powered by NodeBB
                          • Login

                          • Don't have an account? Register

                          • Login or register to search.
                          • First post
                            Last post
                          0
                          • Categories
                          • Recent
                          • Tags
                          • Popular
                          • Users
                          • Groups