0
$\begingroup$

Rosanswers logo

Hi there,

I am new to ROS and would like to know some information about how to go about generating a map using kinect + Pioneer AT Robot? I am not using a tutlebot.

I had earlier generated the map using Laser Scanner. Now, I am trying to generate the map by using Kinect. However, I am stumped as to how to go about with the conversion of pointCloud data espcially since I am not using TurtleBot. Is RGBDSLam one of the ways? I am a little confused here. I could not find much info and hence would be grateful for your help! Thank you!

A small Update, I got the data captured. However, while I am trying to generate the map I keep getting 'Waiting for the Map'. I looked into the frames and noticed that the /map, /camera_link and /camera_depth_frame all were unlinked. Any suggestions as to how to fix this?


Originally posted by Chithra on ROS Answers with karma: 13 on 2013-02-05

Post score: 1


Original comments

Comment by ayush_dewan on 2013-02-06:
For the frame issues use static_tf_publisher to publish transformation between /camera_link and pioneer base_link. The transformation should capture the transformation between position of kinect and base of robot.

$\endgroup$

1 Answer 1

0
$\begingroup$

Rosanswers logo

To use kinect with Pioneer AT u require the usual kinect packages and laserscan_to_pointcloud package. In openni_launch package edit the kinect_frames.launch file. If i remember correctly it defines frames w.r.t to /camera_link which in turn is defined w.r.t to turtlebot by turtlebot package. So u should explicitly define a correct transformation between /camera_link and PionnerAT.

PS: I haven't tried it myself but i feel it should work this way.


Originally posted by ayush_dewan with karma: 1610 on 2013-02-05

This answer was ACCEPTED on the original site

Post score: 1

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.