Hi guys! Thanks for making the source code available. I've been working to run various VIO/VO packages on the datasets collected by our lab quad.
While debugging the problem of divergence of the odometry estimates by VINS-Mono, I found that vins_estimator/odometry topic gives zero angular twist/velocity while running on the entire dataset. I've also checked the mavros/local_position/velocity topic for the angular velocity which is coming out to be non-zero. Even I have found the same situation with the EuRoC MAV Dataset.
Please explain the situation here as I think it is the cause of drifting and scale changes in the odometry results.
Here is the link to the video: https://drive.google.com/open?id=1zDsrJlaXUutUIkD-OjBBlniyNhwRVQjb
Sorry about that. We don't aware we need to publish angular velocity since angular velocity can directly get for raw IMU data. We don't need to publish it again, right?
Hi guys! Thanks for making the source code available. I've been working to run various VIO/VO packages on the datasets collected by our lab quad. While debugging the problem of divergence of the odometry estimates by VINS-Mono, I found that vins_estimator/odometry topic gives zero angular twist/velocity while running on the entire dataset. I've also checked the mavros/local_position/velocity topic for the angular velocity which is coming out to be non-zero. Even I have found the same situation with the EuRoC MAV Dataset. Please explain the situation here as I think it is the cause of drifting and scale changes in the odometry results. Here is the link to the video: https://drive.google.com/open?id=1zDsrJlaXUutUIkD-OjBBlniyNhwRVQjb