Rviz Pointcloud2 stop receiving data when moving fast Kinect.
Hi community,
I'm new to working with Kinect and ROS in general, so I guess i'm missing something here.
I am using ROS Indigo with Ubuntu 14.04 and the Freenect driver.
I show you the steps I make in case you see something irregular:
So i connect the Kinect to ROS:
$ roslaunch freenect_launch freenect.launch depth_registration:=true
I use this RTAB-MAP command to run my Rviz, although I could use, of course, RTAB-MAP. (I do it this way because like this i don't have to configure myself the Rviz for working with the Kinect):
$ roslaunch rtabmap_ros rgbd_mapping.launch rtabmap_args:="--delete_db_on_start" rviz:=true rtabmapviz:=false
Then the Rviz runs normally until i make relatively fast movements with the Kinect (they aren't really fast, just normal human movements), that's when the data streaming stops (so Rviz doesn't refresh) and I receive this kind of notifications via terminal:
...
[ WARN] (2016-04-15 10:28:35.722) OdometryBOW.cpp:305::computeTransform() Not enough correspondences (0 < 20)
[ INFO] [1460708915.727316809]: Odom: quality=0, std dev=0.000000m, update time=0.339010s
[ WARN] [1460708915.831955475]: rtabmap: Could not get transform from odom to camera_link after 0.100000 second!
[ WARN] [1460708915.832098033]: Could not get TF transform from odom to camera_link, sensors will not be synchronized with odometry pose.
[ERROR] (2016-04-15 10:28:35.833) Rtabmap.cpp:890::process() RGB-D SLAM mode is enabled and no odometry is provided. Image 9122 is ignored!
[ INFO] [1460708915.833361176]: rtabmap: Rate=1.00s, Limit=0.000s, RTAB-Map=0.0001s, Pub=0.0000s (local map=162, WM=162)
[ WARN] (2016-04-15 10:28:36.100) OdometryBOW.cpp:305::computeTransform() Not enough correspondences (0 < 20)
[ INFO] [1460708916.105483806]: Odom: quality=0, std dev=0.000000m, update time=0.359968s
....
So this is the main problem, despite that the map that shows in the Rviz is not really accurate and, although it's beautiful to see a 3D map of my living-room, is still some fuzzy itself.
Any thoughts for improving this? Thanks everyone in advance.
Asked by thepirate16 on 2016-04-18 06:59:58 UTC
Comments