ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange
Ask Your Question

Multiple Kinects in Rviz, image fusion?

asked 2013-11-13 06:28:42 -0500

Athoesen gravatar image

updated 2016-10-24 09:00:43 -0500

ngrennan gravatar image

Currently I've got three Kinects (1473's) running on Ubuntu 12.04 with ROSHydro. I'm able to access all image streams from all three Kinects, and the next step I wanted to achieve was combining streams to create a three dimensional (or at least partially 3D) image in something like RViz. I know that I need to create a transform between them but I can't find any tutorials or similar questions on here. Can someone help or point me in the right direction?

edit retag flag offensive close merge delete

1 Answer

Sort by » oldest newest most voted

answered 2013-11-13 06:47:01 -0500

updated 2013-11-25 01:13:46 -0500

The camera_pose_calibration package essentially provides exactly what you´re looking for. Haven´t used it myself, so not sure how mature and easy to adapt it is. The calibration package also is able to estimate transforms, but is originally meant for robot kinematic calibration. Based on it, a approach using the Google Ceres solver is also available. There are of course also other options, like registration using some PCL tools (if you have significant point cloud overlap), using AR markers visible in multiple views and others, which likely require some more work on your part.

edit flag offensive delete link more


I appears that the package does not work straight from getting it but there is a github fork which may resolve problems.

Athoesen gravatar image Athoesen  ( 2014-01-09 09:08:50 -0500 )edit

Question Tools

1 follower


Asked: 2013-11-13 06:28:42 -0500

Seen: 503 times

Last updated: Nov 25 '13