Hi everyone. I am currently looking for an algorithm that will help me provide odometry data for amcl considering that I do not get any odometry from the wheel encoders. In ros1, there was the possibility to feed odometry data from hector mapping to amcl, however in ros humble hector mapping is not yet(or will not be at all) integrated. I am looking for some alternative that can provide somewhat accurate odometry for amcl. You advice will be much appreciated.
Hello @jindecasa,
there are several different ways in which you can get odometry data (local localization) apart from using wheel encoders. These methods require the use of different sensors. Some of these methods are:
- Visual Odometry (uses the camera)
- Inertial Odometry (uses an IMU sensor)
- Laser Odometry (uses the laser scanner)
Each method has advantages and disadvantages. In general, in order to improve odometry it is common to use a sensor fusion package to fuse two or more sources of odometry data.
Hope this helps,
Roberto
Hello @jindecasa ,
If you are looking for the math of sensor fusion algorithms, the Extended Kalman Filter, among many other algorithms, will be a good thing to look into.
Best,
Arunabh
This topic was automatically closed 10 days after the last reply. New replies are no longer allowed.