RGB (640*480) + Depth Image to Real World Kinect [closed]

asked 2015-06-28 21:29:36 -0500

HugoLqsticot gravatar image

Hello, I'm looking for an easy way to Convert kinect RGB and depth values to XYZ coordinates in ROS. My goal is a function with an input of: RGB and depth values of each point of my object, and output of: x,y and z values of each point.

Does anyone have an idea for that ?

edit retag flag offensive reopen merge delete

Closed for the following reason the question is answered, right answer was accepted by HugoLqsticot
close date 2015-06-30 22:51:48.233534


Why don't you use point clouds?

Willson Amalraj gravatar image Willson Amalraj  ( 2015-06-28 23:27:47 -0500 )edit

Because, I recognize my object with colour in openCV in a RGB image, that's my problem. And I think I couldn't do that with a pointCloud

HugoLqsticot gravatar image HugoLqsticot  ( 2015-06-29 21:17:07 -0500 )edit

You can use pin-hole camera model to get x and y co-ordinates if you manage to get z co-ordinate (either using disparity image or something else).

Willson Amalraj gravatar image Willson Amalraj  ( 2015-06-30 05:21:58 -0500 )edit

Thanks! Works Great !

HugoLqsticot gravatar image HugoLqsticot  ( 2015-06-30 22:51:22 -0500 )edit

Hi Hugo, can you please tell me how to get z coordinate? I have u,v from RGB image and I have Depth image . but I don't know what to do now.

PKumars gravatar image PKumars  ( 2016-02-23 06:07:26 -0500 )edit

I want to get exact coordinates of my specific object. I'm subscribing /camera/rgb/image_color and /camera/depth/image_raw. I have u,v pixel coordinate from RGB image. i followed this http://nicolas.burrus.name/index.php/.. . . but couldn't.

PKumars gravatar image PKumars  ( 2016-02-24 06:01:34 -0500 )edit