CathIAS / TLIO

Tight Learned Inertial Odometry
Other
277 stars 69 forks source link

some question about Data Prepare #36

Open Claudbbbbb opened 9 months ago

Claudbbbbb commented 9 months ago

Hi everyong!

I have some questions about the new data file imu_samples_0.csv and imu0_resampled.npy. In my understanding, imu_samples_0.csv contain the original IMU datas, which acc and gyr data were not calibrated, and imu0_resampled.npy's IMU data were calibrated. Could you tell me how original IMU data convert into calibrated IMU data? Just did some gravity alignment? (Rotate the axises of IMU and make the z-axis coincide with the direction of gravitry?) And let the calibrated acc data near [0, 0, 9.8] when the IMU is static?

Thank you for your explanation.

PabloRauhCorro commented 9 months ago

Hello Claudbbbbb,

have you figured it out? Your assumption sounds plausible to me regarding the "rotated" part. But I am unsure what "compensated" is supposed to mean in the context of gyroscope measurements:

{ "columns_name(width)": [ "ts_us(1)", "gyr_compensated_rotated_in_World(3)", "acc_compensated_rotated_in_World(3)", "qxyzw_World_Device(4)", "pos_World_Device(3)", "vel_World(3)" ], "num_rows": 71333, "approximate_frequency_hz": 200.0, "t_start_us": 2851904885.0, "t_end_us": 3208564885.0 }