Starling 2 qvio/odometry covariance all zeroes
-
I read that the "quality" reported by the qvio process is generated from the covariance matrix of the onboard state estimation.
After running voxl_mpa_to_ros, these covariance values are all set to zero.
Is there any work around/configuration for the starling to report the actual covariance values it uses to generate the quality?