The site is read-only. Please transition to use Robotics Stack Exchange
| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
 | 1 | initial version |
I'm not sure what kind of results you can expect when you don't have vehicle odometry and you aren't fusing any absolute yaw measurements, but I guess it doesn't hurt to try. I would start with taking a closer look at the IMU data you're working with.
First, I would make sure that the numbers being reported by your IMU are realistic. Just looking at the bag file, the angular velocities look suspicious, and it looks like your linear accelerations are being reported in g's. If you can, look at the IMU data when the vehicle is stationary and make sure the numbers make sense.
Second, make sure that your IMU data is being reported in the correct ENU frame (i.e., what r_l and navsat_transform are expecting). The details of the IMU co-ordinate frame are outlined in REP-105.
Out of curiosity, what IMU are you working with? If it has a magnetometer you might consider using the madgwick filter or complementary filter to get an orientation estimate.
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
