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

Convert a 16 bit compressed kinect depth image to something usable?

asked 2013-07-18 05:45:28 -0500

madmax gravatar image

updated 2014-01-28 17:17:19 -0500

ngrennan gravatar image


I don't know why, but I have a 16 bit Kinect depth image recorded in my bag file.
And now I want to use rgbd_slam for processing.

Is there a solution to get this working, because the normal bitrate is 32Bit depth, isn't it?

edit retag flag offensive close merge delete

1 Answer

Sort by ยป oldest newest most voted

answered 2013-07-21 01:57:37 -0500

What other data is in your bagfile, please post the output of rosbag info <yourbagfile>. If you have a cloud, set the topic for the depth image in rgbdslam's parameters to "" (empty string). Otherwise, try setting the "depth_scaling" parameter to 0.001 (I'm guessing that the 16bit thing contains millimeters).

I am unsure what you mean by "compressed" in the title? That the bagfile is compressed? That wouldn't have any impact (except the cpu performance required to decompress).

edit flag offensive delete link more


Well, you answered my [other question]( regarding the strange depth image in my bag file. The normal image_transport decompressing does not work for me, so I looked how rqt_gui did the thumbnail view, and took this approach. The just make a python image out of it and I convert it back to a ros image. So now I have this image, but then RGBD throws an opencv error that says it can't convert to bayer_grbg8 or so. Don't know the error exactly, as I am not at university right now...

madmax gravatar image madmax  ( 2013-07-22 07:14:58 -0500 )edit

Question Tools


Asked: 2013-07-18 05:45:28 -0500

Seen: 826 times

Last updated: Jul 21 '13