ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Revision history [back]

click to hide/show revision 1
initial version

Is your question about convert IMU readings to estimate the orientation & position or is it about sending the IMU data using ROS?

For the latter, you can use the ros msg called sensor_msgs/Imu.

As for the former, it is a state estimation problem. You can use standard AHRS algorithms like Mahony/Madgwick filters (link) to convert the sensor readings to estimate the orientation (quaternion). Note, IMU (accelerometer) readings alone will not suffice to estimate the position info accurately, you would need some kind of external feedback like encoders, images (for VIO) etc. But naively, you can integrate the (accelerometer readings - bias) to compute the velocity and position, but these values will drift very quickly no matter what.

Is your question about convert converting IMU readings to estimate the orientation & position or is it about sending the IMU data using ROS?

For the latter, you can use the ros msg called sensor_msgs/Imu.

As for the former, it is a state estimation problem. You can use standard AHRS algorithms like Mahony/Madgwick filters (link) to convert the sensor readings to estimate the orientation (quaternion). Note, IMU (accelerometer) readings alone will not suffice to estimate the position info accurately, you would need some kind of external feedback like encoders, images (for VIO) etc. But naively, you can integrate the (accelerometer readings - bias) to compute the velocity and position, but these values will drift very quickly no matter what. what.