search

daniilidis-group / mvsec

Multi Vehicle Stereo Event Camera Dataset
MIT License
53 stars 10 forks source link

questions about the noise of IMU data and image data #8

Open ghost opened 4 years ago

ghost commented 4 years ago

Hi，First of all, thank you for sharing your dataset. I have some questions about the noise of IMU data. First：I want to use the dataset to run VINS mono，I want to know the paramaters as follows: (1)accelerometer measurement noise standard deviation, (2)gyroscope measurement noise standard deviation, (3)accelerometer bias random work noise standard deviation (4)gyroscope bias random work noise standard deviation and gravity magnitude in your dataset. Second: I found there is no image data in rosbag file, how can I get the image data? Thanks for your time.

alexzzhu commented 4 years ago

Hi there,

The IMU on the DAVIS is the MPU 6150, and you can find the specs sheet here: https://www.cdiweb.com/products/detail/mpu6150-invensense/470090/. This should have the statistics you are looking for. Regarding the image data, which bag are you referring to specifically? It should be available for most of them, although the right camera images are missing in outdoor_day1 and 2.

ghost commented 4 years ago

@alexzzhu Hi,alexzzhu, I would like to know the calibration data of IMU when obtaining the data set, well, Is the device calibrated before data collection? And I am looking for the Indoor flying (Note: No VI-Sensor data is available for this scene)dataset's image file. Thanks a lot!

alexzzhu commented 4 years ago

Oh I see, no we do not have IMU calibrations beyond geometric extrinsics. The images should be stored in the bag file under /davis/image_raw, if that's what you're using. They're also available in the hdf5 files, I believe under the same name.

ghost commented 4 years ago

Please excuse me，calibration is very important for the stability of SLAM/VIO system , I want to know why do not have IMU calibrations beyond geometric extrinsics？

alexzzhu commented 4 years ago

I agree that having the IMU parameters would help with state estimation. However, I'm not sure what to tell you aside from the fact that these were not collected.

ghost commented 4 years ago

I want to use this dataset to run the VINS-mono system,I need the (1)accelerometer measurement noise standard deviation, (2)gyroscope measurement noise standard deviation, (3)accelerometer bias random work noise standard deviation (4)gyroscope bias random work noise standard deviation and gravity magnitude in your dataset. These parameters are very important for the initialization and normal operation of the system, If these parameters are not accurate, I don't know where the IMU can fly?

alexzzhu commented 4 years ago

Have you tried running the system with some default set of calibrations? As far as I'm aware these systems shouldn't be so sensitive as to break completely without calibrated values.

yao-hexiong commented 10 months ago

Hi，First of all, thank you for sharing your dataset. I have some questions about the noise of IMU data. First：I want to use the dataset to run VINS mono，I want to know the paramaters as follows: (1)accelerometer measurement noise standard deviation, (2)gyroscope measurement noise standard deviation, (3)accelerometer bias random work noise standard deviation (4)gyroscope bias random work noise standard deviation and gravity magnitude in your dataset. Second: I found there is no image data in rosbag file, how can I get the image data? Thanks for your time.

I failed to run the indoor flying dataset on vinsmono, how is your yaml file set up?