robot_pose_ekf issue
Hi all,
I have robot that calculates the odometry using the information of the wheels, and what i usually do is publish the odom as a nav_msgs::odometry which has as frame_id "/world" and as child_frame_id "rear_axis"(where all the calculations are made). This setup works with no problem.
Now my problem, I have to combine the information of odometry with the data from an imu(I have an urdf with the right position of the imu->rear_axis), and i was trying to use the robot_pose_ekf to do that.
I made robot_pose_ekf subscribe the information from odometry and from the imu, and I get a tf : /odom->/base_footprint, however i need to connect tf to the previous tf(original).(because I have to acumulate pointcloud while the robot moves) 1.Question: How do I connect these two tf?
2.Question: The frame /base_footprint and /odom(is this like /world frame?) , what does it mean?
3.Question: If the /base_footprint moves with the robot, mabe I can create a static transformation between /base_footprint and my "/rear_axis". (am I completly worng?), to connect the twe tf's tree.
Did I make my self clear?
If anyone could help I'd appreciate.
Perhaps you should look at this question that I asked a few months ago http://answers.ros.org/question/11682/robot_pose_ekf-with-an-external-sensor @Martin_Gunther provided a very clear explanation of all of the frames in robot_pose_ekf
thanks, it was very useful. :D