I use rviz to visualize the image kinect (mounted on PR2) captured, but I found there exists a weird white region, is this a normal situation? If not, does anyone know how to deal with this problem?
And I also want to ask what's the appropriate Fixed Frame & Target Frame to choose?
When I change the Fixed Frame, it became weirder.
Thanks in advance.
Originally posted by Po-Jen Lai on ROS Answers with karma: 1371 on 2013-03-18
Post score: 0
As also can be inferred from your "Robot Model" display showing an Error State, it appears your tf transform tree is not properly working. When this is the case, the transformations between the different links of your robot are not known and all those links are shown as white centered around the origin. Those are very likely what you see backprojected into your Kinect images as "weird white regions". This happens only for the "Camera" display though, as for this, objects of the 3D visualization are projected into the image. If you use the "Image" display instead, you will get the plain Kinect image.
Originally posted by Stefan Kohlbrecher with karma: 24361 on 2013-03-18
This answer was ACCEPTED on the original site
Post score: 5
Comment by Po-Jen Lai on 2013-03-18:
Thanks~your answer exactly solved this problem, it's a conflict between /base_link(for tf and Robot Model) and /camera_link(for camera display).
