how can i use roborealm software (vision for machines) in ros ??

edit

So I've used Roborealm a little about a year ago. I wasn't terribly impressed with it, but it was fairly easy to get something simple going. I think it's Windows only, so you're going to need at least two computers for now (one Windows for Roborealm, one Linux for ROS).

I didn't have much luck with them, but Roborealm has a list of interfaces that it supports. Of these, you should take the closest look at the text sockets API. As long as you can define your output as ASCII, decimal, hex, or binary then you can pipe that over the network and use it as you want in C++, etc.

That's an awful lot of work, so I'd suggest learning OpenCV. You can then be all on one machine without a great number of hacks (so no virtualization). OpenCV has most of the image processing techniques that Roborealm does, or you can at least find a compatible library. A good book to learn OpenCV is Learning OpenCV by the guys who started OpenCV.

If you're using Roborealm for everything (keyboard/joystick driving, AI, etc), then it will be a lot more work to convert to ROS, but I think it's really worth it. RoboRealm is pretty limited and you're also stuck with the languages they have plugins for. Learning ROS will also give you lots of employable skills and a better understanding of how things work. We may also have better support with ROS answers!