Hi,
I am new to rtabmap_ros. I have installed rtabmap_ros for indigo and was following steps mentioned in [http://wiki.ros.org/rtabmap_ros/Tutorials/StereoOutdoorMapping](http://wiki.ros.org/rtabmap_ros/Tutorials/StereoOutdoorMapping) to run the **demo_stereo_outdoor.launch**.
Using the file [stereo_outdoorB.bag](https://docs.google.com/uc?export=download&confirm=5Ptp&id=0B46akLGdg-uaVEs2SGFzT3ZVSU0) the stereo outdoor mapping is done correctly as you can see in [\[0:00 -> 0:52\]](https://www.youtube.com/watch?v=PuLaFdvmbwI). However, if I use a my .bag file, stereo outdoor mapping is not done correctly as can be seen in [\[0:53 -> 1:23\]](https://youtu.be/PuLaFdvmbwI?t=53).
The my .bag file has the same topics as stereo_outdoorB.bag (with the exception of **/tf** that is published through the .launch file). The used .launch file is in: [https://codeshare.io/an4O03](https://codeshare.io/an4O03)
After run a few seconds the map is no longer updated and the following warning is displayed:
[ WARN] [1501502173.552628407, 1487604089.579723953]: rtabmap: Could not get transform from odom to base_footprint after 0.200000 seconds (for stamp=1487604089.150464)!
[ WARN] (2017-07-31 12:56:13.572) OdometryF2F.cpp:203::computeTransform() Registration failed: "Not enough inliers 0/10 (matches=225) between 0 and 0"
You can find the rosbag file that I used in the video at: [sev_2017-02-20-15-20-50.bag](http://vcriis01.inesctec.pt/datasets/DataSet/RailInspect/sev_2017-02-20-15-20-50.bag)
Topics have to be remapped according to the following:
rosbag play --clock sev_2017-02-20-15-20-50.bag /stereo/left/image_raw/compressed:=/stereo_camera/left/image_raw_throttle/compressed /stereo/right/image_raw/compressed:=/stereo_camera/right/image_raw_throttle/compressed /stereo/left/camera_info:=/stereo_camera/left/camera_info_throttle /stereo/right/camera_info:=/stereo_camera/right/camera_info_throttle
Can anyone help me out to solve this?
Best regards,
Jorge Mendes
---
**Edit:** Thanks for the help Mathieu ([@matlabbe](https://answers.ros.org/users/20914/matlabbe/)), but unfortunately I could not get the same result as you. When I run the .launch file that you provided, the following errors and warning appear:
[ERROR] [1504001352.895939351]: Skipping XML Document "/opt/ros/indigo/share/hector_pose_estimation/hector_pose_estimation_nodelets.xml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1504001353.579416634, 1487604051.226342601]: Skipping XML Document "/opt/ros/indigo/share/hector_pose_estimation/hector_pose_estimation_nodelets.xml" which had no Root Element. This likely means the XML is malformed or missing.
[ERROR] [1504001353.621420049]: Skipping XML Document "/opt/ros/indigo/share/hector_pose_estimation/hector_pose_estimation_nodelets.xml" which had no Root Element. This likely means the XML is malformed or missing.
>
[ WARN] (2017-08-29 11:09:20.416) Features2d.cpp:425::create() SURF/SIFT features cannot be used because OpenCV was not built with nonfree module. ORB is used instead.
You can see the result of the execution on my computer in the following video: [RTAB-Map ROS stereo outdoor mapping problem](https://youtu.be/Te5myteDM5k)
When I play the .bag file continues to appear warnings of the type:
[ WARN] [1504001428.405463965, 1487604078.162979112]: rtabmap: Could not get transform from odom to base_footprint after 0.200000 seconds (for stamp=1487604077.691293)!
[ WARN] [1504001428.435126342, 1487604078.193218747]: rtabmapviz: Could not get transform from odom to base_footprint after 0,200000 seconds (for stamp=1487604077,491246)!
[ WARN] [1504001428.606564760, 1487604078.364592901]: rtabmap: Could not get transform from odom to base_footprint after 0.200000 seconds (for stamp=1487604077.891221)!
[ WARN] (2017-08-29 11:10:28.657) OdometryF2M.cpp:228::computeTransform() Registration failed: "Not enough inliers 0/10 (matches=93) between -1 and 0"
and
[ WARN] (2017-08-29 11:10:30.108) Stereo.cpp:159::computeCorrespondences() A large number (536/925) of stereo correspondences are rejected! Optical flow may have failed, images are not calibrated, the background is too far (no disparity between the images) or maximum disparity may be too small (64).
Do you have any idea of which be the reason for not getting the same result as you? Did I do something wrong?
Thanks,
Jorge
↧