Open tonycar12002 opened 5 years ago
See Odometry Lost section on this page: https://github.com/introlab/rtabmap/wiki/Kinect-mapping#lostodometry
You could start by using the wheel odometry from the husky as the input odometry for rtabmap. Then afterwards, if not accurate enough, you may check for lidar odometry with velodyne and/or visual odometry from RGB-D camera, and do fusion with robot_localization package.
[ WARN] (2019-03-02 11:36:50.149) OdometryF2M.cpp:469::computeTransform() Registration failed: "Not enough inliers 0/20 (matches=2) between -1 and 1141" [ INFO] [1551497810.151549105, 39055.260000000]: Odom: quality=0, std dev=0.000000m|0.000000rad, update time=0.052208s