Hey all,

I'm in the process of trying to perform rgbdslam with kinect2 data. I've recorded a bag file with rgb/image depth/image and camera info (generated by calibration). My setup is ros hydro with ubuntu 12.04.

When I'm playing back the bag file, after modifying the topics in the launch file, I get the following error:

[ERROR] [1426850809.548410517]: depth and visual image differ in size! Ignoring Data
...
[ERROR] [1426850809.548410517]: depth and visual image differ in size! Ignoring Data

My guess is that since OpenNi publishes same size rgb/depth frames generated by the kinect sensor, the kinect2 sensors' non equal sizes cause a miss-match. Has anyone tried to overcome this? I can upload my .bag file in case anyone wants to reproduce this.

Thanks in advance.