coordinate projections from camera to workspace

Hello all,

My project is to design a pick and place robotic arm with the help of 2D image processing. Any idea how can I obtain the real world coordinates with respect to pixel coordinates used in the image processing algorithms? Because I have to tell the robotic arm the real coordinates if I were to pick those objects.

I have found a few related cores that might be useful for this purpose. For instance: ofcamera or ofVec3f. Advice needed. Thank you so much in advanced.

you can’t with just one camera, you’ll be unable to get the world coordinates.

Check out the chessboard technique or something similar.
see this:
it means your object has to be a chessboard.

Or two cameras and stereo calibration. (google for it, some ofx plugins/demos avialable)

Why not use 3D camera’s like the kinect? They give you exactly what you need.