Ezio46
commented
4 months ago Can you please also tell me why the scale of the trajectory obtained in RVIZ from dm-vio (I make it from pose pointcloud) is several times smaller than the scale of the trajectory in Rviz obtained from Ros pose
Hello, I get this error when adding vignette:
rosrun dmvio_ros node nogui=0 useimu=1 quiet=1 mode=3 calib=/home/igor/RealSense/Final/camera1.txt imuCalib=/home/igor/RealSense/Final/camchain.yaml gamma=/home/igor/RealSense/Final/pcalib.txt vignette=/home/igor/RealSense/Final/vignette_t2652.png settingsFile=/home/igor/RealSense/Final/t265_noise_tumvi.yaml cam0/image_raw:=/iris/c920/image_raw imu0:=/imu Enabling IMU integration! QUIET MODE, I'll shut up! PHOTOMETRIC MODE WITH CALIBRATION, BUT NO OR INACCURATE EXPOSURE! Loading settings from yaml file: /home/igor/RealSense/Final/t265_noise_tumvi.yaml! Settings: accelerometer_noise_density: 0.224 accelerometer_random_walk: 0.043 addVisualToCoarseGraphIfTrackingBad: 0 alwaysCanBreakIMU: 0 baToCoarseAccBiasVariance: 1000 baToCoarseGyrBiasVariance: 0.05 baToCoarsePoseVariance: 0.1 baToCoarseRotVariance: 1 baToCoarseVelVariance: 0.1 calib: /home/igor/RealSense/Final/camera1.txt dynamicWeightRMSEThresh: 8 fixKeyframeDuringCoarseTracking: 1 gamma: /home/igor/RealSense/Final/pcalib.txt generalScaleIntervalSize: 60 gravityDirectionFixZ: 1 gyroscope_noise_density: 0.0128 gyroscope_random_walk: 0.0011 imuCalib: /home/igor/RealSense/Final/camchain.yaml init_coarseScaleUncertaintyThresh: 5 init_conversionType: 0 init_disableVIOUntilFirstInit: 1 init_fixPoses: 1 init_initDSOParams: 1 init_lambdaLowerBound: 1e-16 init_maxNumPoses: 100 init_multipleBiases: 0 init_multithreadedInitDespiteNonRT: 0 init_onlyKFs: 1 init_pgba_conversionType: 0 init_pgba_delay: 100 init_pgba_prepareGraphAddDelValues: 0 init_pgba_prepareGraphAddFactors: 0 init_pgba_priorExtrinsicsRot: 0.01 init_pgba_priorExtrinsicsTrans: 0.1 init_pgba_priorGravityDirection: 0.4 init_pgba_priorGravityDirectionZ: 0.0001 init_pgba_reinitScaleUncertaintyThresh: 1 init_pgba_scaleUncertaintyThresh: 5 init_pgba_skipFirstKFs: 1 init_priorExtrinsicsRot: 0.01 init_priorExtrinsicsTrans: 0.1 init_priorGravityDirection: 0.4 init_priorGravityDirectionZ: 0.0001 init_priorRotSigma: 1e-05 init_priorTransSigma: 1e-05 init_requestFullResetErrorThreshold: -1 init_requestFullResetNormalizedErrorThreshold: 2 init_scalePriorAfterInit: 0 init_secondthreshGravdir: 1000 init_secondthreshScale: 1e+07 init_threshGravdir: 1000 init_threshScale: 1.02 init_transitionModel: 2 init_updatePoses: 1 integration_sigma: 0.316227 loadRosbagThread: 0 maxFrameEnergyThreshold: 5000 maxSkipFramesFullReset: -1 maxSkipFramesVisualInertial: 2 maxSkipFramesVisualInit: 0 maxSkipFramesVisualOnlyMode: 1 maxTimeBetweenInitFrames: 1 minQueueSizeForSkipping: 2 normalizeCamSize: 0.2 numMeasurementsGravityInit: 40 preload: 0 priorExtrinsicsRot: 0.01 priorExtrinsicsTrans: 0.1 priorGravityDirection: 0.4 priorGravityDirectionZ: 0.0001 resultsPrefix: rosbag: setting_forceNoKFTranslationThresh: 0 setting_maxOptIterations: 6 setting_minFramesBetweenKeyframes: -0.5 setting_minIdepth: 0.02 setting_minOptIterations: 1 setting_optGravity: 1 setting_optIMUExtrinsics: 0 setting_optScaleBA: 1 setting_prior_bias: 0 setting_prior_velocity: 0 setting_scaleFixTH: 0 setting_solverMode: 2048 setting_transferCovToCoarse: 1 setting_visualOnlyAfterScaleFixing: 0 setting_weightDSOCoarse: 0.001 setting_weightDSOToGTSAM: 1.66667e-05 setting_weightZeroPriorDSOInitX: 0 setting_weightZeroPriorDSOInitY: 0 skipFirstKeyframe: 0 skipFramesVisualOnlyDelay: 30 speed: 0 start: 2 timeshift: 0 transferCovToCoarseMultiplier: 1 updateDynamicWeightDuringOptimization: 1 useScaleDiagonalHack: 0 vignette: /home/igor/RealSense/Final/vignette_t2652.png Reading Calibration from file /home/igor/RealSense/Final/camera1.txt ... found! Input resolution: 1920 1080 In: 1360.470496 1360.470496 960.500000 540.500000 0.000000 Out: Rectify Crop Output resolution: 512 512 finding CROP optimal new model! initial range: x: -0.7131 - 0.7115; y: -0.4012 - 0.3998! iteration 00001: range: x: -0.7095 - 0.7080; y: -0.4012 - 0.3998! iteration 00002: range: x: -0.7059 - 0.7044; y: -0.4012 - 0.3998! iteration 00003: range: x: -0.7059 - 0.7044; y: -0.3992 - 0.3978! iteration 00004: range: x: -0.7059 - 0.7044; y: -0.3972 - 0.3958! iteration 00005: range: x: -0.7059 - 0.7044; y: -0.3972 - 0.3958!
Rectified Kamera Matrix: 362.31 0 255.772 0 644.437 255.951 0 0 1
Reading Photometric Calibration from file /home/igor/RealSense/Final/pcalib.txt Reading Vignette Image from /home/igor/RealSense/Final/vignette_t2652.png Successfully read photometric calibration! using pyramid levels 0 to 3. coarsest resolution: 64 x 64! Loading IMU parameter file at: /home/igor/RealSense/Final/camchain.yaml Used T_cam_imu: -0.999525 0.0296153 -0.00852233 0.0472799 0.00750192 -0.0343974 -0.99938 -0.0474432 -0.0298901 -0.998969 0.0341589 -0.0682 0 0 0 1 Used noise values: 0.043 0.0011 0.224 0.0128 Image topic: /iris/c920/image_raw IMU topic: /imu START PANGOLIN! Switching to initializer state: RealtimeCoarseIMUInitState PixelSelector: Using block sizes: 16, 16 WARNING: Not sending frame, because it does not have IMU data yet. PixelSelector: Using block sizes: 16, 16 node: /home/igor/dm-vio/src/dso/FullSystem/CoarseInitializer.cpp:1062: void dso::CoarseInitializer::makeNN(): Assertion `ret_index[0]>=0 && ret_index[0] < numPoints[lvl+1]' failed. Aborted (core dumped)