How do I compute precision and recall from a dataset? [closed]

edit

I am doing a project on evaluation of different descriptors for stereo image matching, eg. SIFT, SURF,... So far I've got the code to find the features in both images and match them using the tutorials provided. But to evaluate these results, I need some statistics like recall and precision, the problem is I don't know how to verify that these matches are correct. I've read that you need some kind of ground truth, but i don't know where to find it and implement it in OpenCV.

I am fairly new to this so any help would be appreciated!