I think what you are asking is how to make sure that you acquire at 60 fps by having the edge detection occur in parallel with the acquisition. Attached is a way to do this. You create another image ref and copy new data to this one each time you get a new image from the camera, the next time through the loop it will perform the edge detection while a new image is being acquired. As long as the edge detection takes a shorter time than the image acquisition, you will be fine. Otherwise, you may need a more complicated architecture.
Is this what you are asking about?
Randall Pursley