做啥类型网站,营销型网站的评价标准,化工行业网站,博客网站建设设计报告前言
周末休息了两天#xff0c;接着做上周五那个VIRAL数据集没有运行成功的工作。现在的最新OpenVINS需要重新写配置文件#xff0c;不像之前那样都写在launch里#xff0c;因此需要根据数据集情况配置好estimator_config.yaml还有两个标定参数文件。
VIRAL数据集
VIRAL…前言
周末休息了两天接着做上周五那个VIRAL数据集没有运行成功的工作。现在的最新OpenVINS需要重新写配置文件不像之前那样都写在launch里因此需要根据数据集情况配置好estimator_config.yaml还有两个标定参数文件。
VIRAL数据集
VIRAL数据集包含雷达、相机、IMU、UWB四种数据是南洋理工大学在22年发布的。 官网地址https://ntu-aris.github.io/ntu_viral_dataset/ 适配VIRAL的OpenVINS旧版https://github.com/brytsknguyen/open_vins.git VIRAL数据集本身作者对一些常用VIO开源代码做了适配修改其中就包括OpenVINS但是这个是更新之前的OpenVINS现在的使用方式配置和之前有所不同。我刚开始从Euroc的数据集配置改动只是改VIRAL以前OpenVINS配置的参数初始化跑不通如下图所示。 这是VIRAL适配的openvins的配置情况是通过launch进行配置的。
launchparam name/use_sim_time valuetrue /arg namepublish_clock default--clock/!-- NTU VIRAL dataset --!-- EEE --arg namebag_file default/home/merlincs/workspace/dataset/VIRAL/eee_01/eee_01.bag/!-- MASTER NODE! --node namerun_serial_msckf pkgov_msckf typerun_serial_msckf outputscreen clear_paramstrue requiredtrue!-- bag topics --param nametopic_imu typestring value/imu/imu /param nametopic_camera0 typestring value/right/image_raw /param nametopic_camera1 typestring value/left/image_raw /rosparam paramstereo_pairs[0,1]/rosparam!-- bag parameters --param namepath_bag typestring value$(arg bag_file) /!-- param namepath_gt typestring value$(find ov_data)/euroc_mav/V1_01_easy.csv / --!-- param namebag_start typedouble value0 / --!-- param namebag_durr typeint value-1 / --!-- world/filter parameters --param nameuse_fej typebool valuetrue /param nameuse_imuavg typebool valuetrue /param nameuse_rk4int typebool valuetrue /param nameuse_stereo typebool valuetrue /param namecalib_cam_extrinsics typebool valuetrue /param namecalib_cam_intrinsics typebool valuetrue /param namecalib_cam_timeoffset typebool valuetrue /param namecalib_camimu_dt typedouble value0.0 /param namemax_clones typeint value11 /param namemax_slam typeint value75 /param namemax_slam_in_update typeint value25 / !-- 25 seems to work well --param namemax_msckf_in_update typeint value40 /param namemax_cameras typeint value2 /param namedt_slam_delay typedouble value3 /param nameinit_window_time typedouble value0.75 /param nameinit_imu_thresh typedouble value0.25 /rosparam paramgravity[0.0,0.0,9.81]/rosparamparam namefeat_rep_msckf typestring valueGLOBAL_3D /param namefeat_rep_slam typestring valueANCHORED_FULL_INVERSE_DEPTH /param namefeat_rep_aruco typestring valueANCHORED_FULL_INVERSE_DEPTH /!-- zero velocity update parameters --param nametry_zupt typebool valuefalse /param namezupt_chi2_multipler typeint value2 /param namezupt_max_velocity typedouble value0.3 /param namezupt_noise_multiplier typedouble value50 /!-- timing statistics recording --param namerecord_timing_information typebool valuefalse /param namerecord_timing_filepath typestring value/tmp/timing_stereo.txt /!-- tracker/extractor properties --param nameuse_klt typebool valuetrue /param namenum_pts typeint value250 /param namefast_threshold typeint value15 /param namegrid_x typeint value5 /param namegrid_y typeint value3 /param namemin_px_dist typeint value5 /param nameknn_ratio typedouble value0.70 /param namedownsample_cameras typebool valuefalse /param namemulti_threading typebool valuetrue /!-- aruco tag/mapping properties --param nameuse_aruco typebool valuefalse /param namenum_aruco typeint value1024 /param namedownsize_aruco typebool valuetrue /!-- sensor noise values / update --param nameup_msckf_sigma_px typedouble value1 /param nameup_msckf_chi2_multipler typedouble value1 /param nameup_slam_sigma_px typedouble value1 /param nameup_slam_chi2_multipler typedouble value1 /param nameup_aruco_sigma_px typedouble value1 /param nameup_aruco_chi2_multipler typedouble value1 /param namegyroscope_noise_density typedouble value5.0e-3 /param namegyroscope_random_walk typedouble value3.0e-6 /param nameaccelerometer_noise_density typedouble value6.0e-2 /param nameaccelerometer_random_walk typedouble value8.0e-5 /!-- camera intrinsics --rosparam paramcam0_wh[752, 480]/rosparamrosparam paramcam1_wh[752, 480]/rosparamparam namecam0_is_fisheye typebool valuefalse /param namecam1_is_fisheye typebool valuefalse /rosparam paramcam0_k[4.313364265799752e02, 4.327527965378035e02, 3.548956286992647e02, 2.325508916495161e02]/rosparamrosparam paramcam0_d[-0.300267420221178, 0.090544063693053, 3.330220891093334e-05, 8.989607188457415e-05]/rosparamrosparam paramcam1_k[4.250258563372763e02, 4.267976260903337e02, 3.860151866550880e02, 2.419130336743440e02]/rosparamrosparam paramcam1_d[-0.288105327549552, 0.074578284234601, 7.784489598138802e-04, -2.277853975035461e-04]/rosparam!-- camera extrinsics --rosparam paramT_C0toI[-0.01916508, -0.01496218, 0.99970437, 0.00519443,0.99974371, 0.01176483, 0.01934191, 0.1347802,-0.01205075, 0.99981884, 0.01473287, 0.01465067,0.00000000, 0.00000000, 0.00000000, 1.00000000]/rosparamrosparam paramT_C1toI[0.02183084, -0.01312053, 0.99967558, 0.00552943,0.99975965, 0.00230088, -0.02180248, -0.12431302,-0.00201407, 0.99991127, 0.01316761, 0.01614686, 0.00000000, 0.00000000, 0.00000000, 1.00000000]/rosparam/nodenode pkgrviz typerviz nameov_msckf_rviz respawntrue outputlogargs-d $(find ov_msckf)/launch/ntuviral.rviz /!-- arg nameautorun defaultfalse/node required$(arg autorun) pkgrosbag typeplay namerosbag_playargs$(arg publish_clock) $(arg bag_file) -r 1/ --/launch对应把上面参数写入新建的config/viral中三个配置文件后跑不通 主要原因是因为静态初始化运动检测的原因具体原理我也还不是很清楚下一次博客对于初始化这块做详细的学习。因此除了抄viral适配openvins中的配置外还需要对配置文件进行一些改动下面介绍一下配置文件各个参数含义。
配置文件详解
config文件夹内有三个配置文件 estimator_config.yamlkalibr_imucam_chain.yamlkalibr_imu_chain.yaml。 第一个是针对不同数据集对估计器的配置第二个第三个是相机和IMU的标定参数。 下面是针对viral数据集进行修改过的配置文件。目前还只是对eee01.bag这一个数据包初始化有效
1、estimator_config.yaml
%YAML:1.0 # need to specify the file type at the top!verbosity: INFO # ALL, DEBUG, INFO, WARNING, ERROR, SILENTuse_fej: true # if first-estimate Jacobians should be used (enable for good consistency)
integration: rk4 # discrete, rk4, analytical (if rk4 or analytical used then analytical covariance propagation is used)
use_stereo: true # if we have more than 1 camera, if we should try to track stereo constraints between pairs
max_cameras: 2 # how many cameras we have 1 mono, 2 stereo, 2 binocular (all mono tracking)calib_cam_extrinsics: true # if the transform between camera and IMU should be optimized R_ItoC, p_CinI
calib_cam_intrinsics: true # if camera intrinsics should be optimized (focal, center, distortion)
calib_cam_timeoffset: true # if timeoffset between camera and IMU should be optimized
calib_imu_intrinsics: false # if imu intrinsics should be calibrated (rotation and skew-scale matrix)
calib_imu_g_sensitivity: false # if gyroscope gravity sensitivity (Tg) should be calibratedmax_clones: 11 # how many clones in the sliding window
max_slam: 75 # number of features in our state vector
max_slam_in_update: 25 # update can be split into sequential updates of batches, how many in a batch
max_msckf_in_update: 40 # how many MSCKF features to use in the update
dt_slam_delay: 3 # delay before initializing (helps with stability from bad initialization...)gravity_mag: 9.81 # magnitude of gravity in this locationfeat_rep_msckf: GLOBAL_3D
feat_rep_slam: ANCHORED_FULL_INVERSE_DEPTH
feat_rep_aruco: ANCHORED_FULL_INVERSE_DEPTH# zero velocity update parameters we can use
# we support either IMU-based or disparity detection.
try_zupt: false
zupt_chi2_multipler: 2 # set to 0 for only disp-based
zupt_max_velocity: 0.3
zupt_noise_multiplier: 50
zupt_max_disparity: 0.5 # set to 0 for only imu-based
zupt_only_at_beginning: false#
# init_window_time: 0.75 # how many seconds to collect initialization information
init_imu_thresh: 0.25 # threshold for variance of the accelerometer to detect a jerk in motion
init_max_disparity: 1.0 # max disparity to consider the platform stationary (dependent on resolution)
init_max_features: 20 # how many features to track during initialization (saves on computation)init_dyn_use: false # if dynamic initialization should be used
init_dyn_mle_opt_calib: false # if we should optimize calibration during intialization (not recommended)
init_dyn_mle_max_iter: 50 # how many iterations the MLE refinement should use (zero to skip the MLE)
init_dyn_mle_max_time: 0.05 # how many seconds the MLE should be completed in
init_dyn_mle_max_threads: 6 # how many threads the MLE should use
init_dyn_num_pose: 6 # number of poses to use within our window time (evenly spaced)
init_dyn_min_deg: 10.0 # orientation change needed to try to initinit_dyn_inflation_ori: 10 # what to inflate the recovered q_GtoI covariance by
init_dyn_inflation_vel: 100 # what to inflate the recovered v_IinG covariance by
init_dyn_inflation_bg: 10 # what to inflate the recovered bias_g covariance by
init_dyn_inflation_ba: 100 # what to inflate the recovered bias_a covariance by
init_dyn_min_rec_cond: 1e-12 # reciprocal condition number thresh for info inversioninit_dyn_bias_g: [ 0.0, 0.0, 0.0 ] # initial gyroscope bias guess
init_dyn_bias_a: [ 0.0, 0.0, 0.0 ] # initial accelerometer bias guess#
# record_timing_information: false # if we want to record timing information of the method
record_timing_filepath: /tmp/traj_timing.txt # https://docs.openvins.com/eval-timing.html#eval-ov-timing-flame# if we want to save the simulation state and its diagional covariance
# use this with rosrun ov_eval error_simulation
save_total_state: false
filepath_est: /tmp/ov_estimate.txt
filepath_std: /tmp/ov_estimate_std.txt
filepath_gt: /tmp/ov_groundtruth.txt#
# # our front-end feature tracking parameters
# we have a KLT and descriptor based (KLT is better implemented...)
use_klt: true # if true we will use KLT, otherwise use a ORB descriptor robust matching
num_pts: 250 # number of points (per camera) we will extract and try to track
fast_threshold: 15 # threshold for fast extraction (warning: lower threshs can be expensive)
grid_x: 5 # extraction sub-grid count for horizontal direction (uniform tracking)
grid_y: 3 # extraction sub-grid count for vertical direction (uniform tracking)
min_px_dist: 5 # distance between features (features near each other provide less information)
knn_ratio: 0.70 # descriptor knn threshold for the top two descriptor matches
track_frequency: 11.0 # frequency we will perform feature tracking at (in frames per second / hertz)
downsample_cameras: false # will downsample image in half if true
num_opencv_threads: -1 # -1: auto, 0-1: serial, 1: number of threads
histogram_method: HISTOGRAM # NONE, HISTOGRAM, CLAHE# aruco tag tracker for the system
# DICT_6X6_1000 from https://chev.me/arucogen/
use_aruco: false
num_aruco: 1024
downsize_aruco: true#
# # camera noises and chi-squared threshold multipliers
up_msckf_sigma_px: 1
up_msckf_chi2_multipler: 1
up_slam_sigma_px: 1
up_slam_chi2_multipler: 1
up_aruco_sigma_px: 1
up_aruco_chi2_multipler: 1# masks for our images
use_mask: false# imu and camera spacial-temporal
# imu config should also have the correct noise values
relative_config_imu: kalibr_imu_chain.yaml
relative_config_imucam: kalibr_imucam_chain.yaml2、kalibr_imucam_chain.yaml
%YAML:1.0cam0:T_imu_cam: #rotation from camera to IMU R_CtoI, position of camera in IMU p_CinI- [-0.01916508, -0.01496218, 0.99970437, 0.00519443]- [0.99974371, 0.01176483, 0.01934191, 0.1347802]- [-0.01205075, 0.99981884, 0.01473287, 0.01465067]- [0.0, 0.0, 0.0, 1.0]cam_overlaps: [1]camera_model: pinhole#相机模型distortion_coeffs: [-0.300267420221178, 0.090544063693053, 3.330220891093334e-05, 8.989607188457415e-05]#畸变参数distortion_model: radtan#畸变模型intrinsics: [4.313364265799752e02, 4.327527965378035e02, 3.548956286992647e02, 2.325508916495161e02] #fu, fv, cu, cvresolution: [752, 480]#分辨率rostopic: /right/image_raw
cam1:T_imu_cam: #rotation from camera to IMU R_CtoI, position of camera in IMU p_CinI- [0.02183084, -0.01312053, 0.99967558, 0.00552943]- [0.99975965, 0.00230088, -0.02180248, -0.12431302]- [-0.00201407, 0.99991127, 0.01316761, 0.01614686]- [0.0, 0.0, 0.0, 1.0]cam_overlaps: [0]camera_model: pinholedistortion_coeffs: [-0.288105327549552, 0.074578284234601, 7.784489598138802e-04, -2.277853975035461e-04]distortion_model: radtanintrinsics: [4.250258563372763e02, 4.267976260903337e02, 3.860151866550880e02, 2.419130336743440e02] #fu, fv, cu, cvresolution: [752, 480]rostopic: /left/image_raw3、kalibr_imu_chain.yaml
%YAML:1.0imu0:T_i_b:- [1.0, 0.0, 0.0, 0.0]- [0.0, 1.0, 0.0, 0.0]- [0.0, 0.0, 1.0, 0.0]- [0.0, 0.0, 0.0, 1.0]accelerometer_noise_density: 6.0e-2 # [ m / s^2 / sqrt(Hz) ] ( accel white noise )accelerometer_random_walk: 8.0e-5 # [ m / s^3 / sqrt(Hz) ]. ( accel bias diffusion )gyroscope_noise_density: 5.0e-3 # [ rad / s / sqrt(Hz) ] ( gyro white noise )gyroscope_random_walk: 3.0e-6 # [ rad / s^2 / sqrt(Hz) ] ( gyro bias diffusion )rostopic: /imu/imutime_offset: 0.0update_rate: 385.0#IMU更新频率# three different modes supported:# calibrated (same as kalibr), kalibr, rpngmodel: kalibr# how to get from Kalibr imu.yaml result file:# - Tw is imu0:gyroscopes:M:# - R_IMUtoGYRO: is imu0:gyroscopes:C_gyro_i:# - Ta is imu0:accelerometers:M:# - R_IMUtoACC not used by Kalibr# - Tg is imu0:gyroscopes:A:Tw:- [ 1.0, 0.0, 0.0 ]- [ 0.0, 1.0, 0.0 ]- [ 0.0, 0.0, 1.0 ]R_IMUtoGYRO:- [ 1.0, 0.0, 0.0 ]- [ 0.0, 1.0, 0.0 ]- [ 0.0, 0.0, 1.0 ]Ta:- [ 1.0, 0.0, 0.0 ]- [ 0.0, 1.0, 0.0 ]- [ 0.0, 0.0, 1.0 ]R_IMUtoACC:- [ 1.0, 0.0, 0.0 ]- [ 0.0, 1.0, 0.0 ]- [ 0.0, 0.0, 1.0 ]Tg:- [ 0.0, 0.0, 0.0 ]- [ 0.0, 0.0, 0.0 ]- [ 0.0, 0.0, 0.0 ]
实验结果
按照上面进行配置文件修改然后运行如下命令
#第一个终端
roscore#第二个终端
source devel/setup.bash
roslaunch ov_msckf subscribe.launch config:viral#第三个终端
rviz
#然后导入配置ntuviral.rviz从viral适配的openvins中下载在ov_msckf/launch中#数据文件夹下打开第四个终端
rosbag play eee_01.bag
运行结果如图所示 现在还只能在eee01.bag这一个数据包初始化能跑通同样的配置跑eee02.bag就不行初始化这块还是要明白原理才能够更好地进行配置。接下来重点学习一下OpenVINS的初始化原理看看怎么配置静态初始化和动态初始化新版本开源的新功能应该很好用。
