Computer vision in ROS [closed]

edit

I am new in computer vision but I need to implement a robot with these kind of abilities.

1. It has to identify objects with a particular color and size.
2. For this purpose I have a RGB camera which gives me a topic of the type: sensor_msgs/Image

I imagine that ROS has its own version of OpenCV for these cases. What I have to do is a pretty simple task but I have no idea of image processing.

Any idea? Where can I find libraries and documentation explaining them?

(I am working with ROS Hydro)

Thank you