The site is read-only. Please transition to use Robotics Stack Exchange
| ROS Resources: Documentation | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange |
 | 1 | initial version |
Hi there,
You don't need to convert the data on the depth image topic manually. If you execute roslaunch openni_launch openni.launch you should have the topic /camera/depth/points which is a 3D Point Cloud with the depth image converted to what you want.
About OpenCV, unless you need the bleeding edge functionalities of the ultra-super-last version, the one that comes with ROS is fairly recent and should include everything you need.
Regarding the black jpg image, the data on the topic /camera/depth/image contains uint16 depths in mm, which is not a regular image that can be directly saved as jpg, you would need to convert the depth values to the range of a uchar (consequently loosing information, because uchar is only 8bits) to store it as an image.
Also, the best way to get started with opencv is to follow the tutorials and ask in answers.opencv.org all your questions about the computer vision library.
I hope this helps
ROS Answers is licensed under Creative Commons Attribution 3.0 Content on this site is licensed under a Creative Commons Attribution Share Alike 3.0 license.
