Software Archive
Read-only legacy content
17061 Discussions

Intel RealSense Basics


I recently have taken on the task of attaching a Intel RealSense R200 3D camera to a UR5 6axis Robot Arm with the goal of giving the robot arm image processing abilities. Unfortunately, I am having very basic issues getting started with the Intel RealSense Camera and hope some of you may have some documentation to help clear this up. On the intel site it says a lot about what the camera can do, but there is little advice on how anything is done. I am having big problems getting the camera booted up, getting the camera to interface with my computer, and getting to a state where programs can be run on the camera. 


One confusing part about the camera is its lack of user interface (no buttons, switches or anything) Here are my very basic questions... 


1) How to operate this device? (I mean literally I haven't even gotten it turned on or doing anything) 


2) What software would I need to download for my computer to recognize the device? (I connect the RealSense to a computer and the computer doesn't even recognize its there) 


3)  What are the libraries out there for programming this device? 


4) What is the best way to setup communication between the camera and a robot arm? 

0 Kudos
1 Reply
Valued Contributor II

1) It's a USB device. When you plug it in to a USB3 slot, it's on.

2) You need to download and install DCM (Where it says R200 Camera Driver):

3) There are 2:

RealSense SDK provides acquisition and computer vision functionalities such as projections, face detection, hand tracking, etc. It is Windows only, and limited to some CPUs.

librealsense provides acquisition and stream projections. This runs on Windows, Linux, Mac. It doesn't provide computer vision functionalities, but you can use third party tools such as OpenCV, PCL, VTK, etc.

In your case, I would recommend librealsense.

4) Get a computer with USB3 and digital I/O. Some suggestions would be the Minnowboard Max board, or the odroid. Or you can just have a PC for camera acquisition, and then connect to some low level board through the network.

0 Kudos